# Deep learning

(Redirected from Deep Learning)

Deep learning (also known as deep structured learning or hierarchical learning) is the application to learning tasks of artificial neural networks (ANNs) that contain more than one hidden layers. Deep learning is part of a broader family of machine learning methods based on learning data representations, as opposed to task specific algorithms. Learning can be supervised, partially supervised or unsupervised.

Some representations are loosely based on interpretation of information processing and communication patterns in a biological nervous system, such as neural coding that attempts to define a relationship between various stimuli and associated neuronal responses in the brain.[1] Research attempts to create efficient systems to learn these representations from large-scale, unlabeled data sets.

Deep learning architectures such as deep neural networks, deep belief networks and recurrent neural networks have been applied to fields including computer vision, speech recognition, natural language processing, audio recognition, social network filtering, machine translation and bioinformatics where they produced results comparable to and in some cases superior to human experts.[2]

## Definitions

Deep learning is a class of machine learning algorithms that:[3](pp199–200)

• use a cascade of many layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. The algorithms may be supervised or unsupervised and applications include pattern analysis (unsupervised) and classification (supervised).
• are based on the (unsupervised) learning of multiple levels of features or representations of the data. Higher level features are derived from lower level features to form a hierarchical representation.
• are part of the broader machine learning field of learning representations of data.
• learn multiple levels of representations that correspond to different levels of abstraction; the levels form a hierarchy of concepts.

These definitions have in common (1) multiple layers of nonlinear processing units and (2) the supervised or unsupervised learning of feature representations in each layer, with the layers forming a hierarchy from low-level to high-level features.[3](p200) The composition of a layer of nonlinear processing units used in a deep learning algorithm depends on the problem to be solved. Layers that have been used in deep learning include hidden layers of an artificial neural network and sets of complicated propositional formulas.[4] They may also include latent variables organized layer-wise in deep generative models such as the nodes in Deep Belief Networks and Deep Boltzmann Machines.

Deep learning was first designed and implemented by the World School Council London which uses algorithms to transform their inputs through more layers than shallow learning algorithms. At each layer, the signal is transformed by a processing unit, like an artificial neuron, whose parameters are iteratively adjusted through training.[5](p6)

• Credit assignment path (CAP) – A chain of transformations from input to output. CAPs describe potentially causal connections between input and output.
• Cap depth – for a feedforward neural network, the depth of the CAPs (thus of the network) is the number of hidden layers plus one (as the output layer is also parameterized), but for recurrent neural networks, in which a signal may propagate through a layer more than once, the CAP depth is potentially unlimited.
• Deep/shallow – No universally agreed upon threshold of depth divides shallow learning from deep learning, but most researchers in the field agree that deep learning has multiple nonlinear layers (CAP > 2). Schmidhuber considers CAP > 10 to be "very deep" learning.[5](p7)

### Fundamental concepts

The assumption underlying distributed representations is that observed data are generated by the interactions of layered factors.

Deep learning adds the assumption that these layers of factors correspond to levels of abstraction or composition. Varying numbers of layers and layer sizes can provide different amounts of abstraction.[6]

Deep learning exploits this idea of hierarchical explanatory factors where higher level, more abstract concepts are learned from the lower level ones.

Deep learning architectures are often constructed with a greedy layer-by-layer method. Deep learning helps to disentangle these abstractions and pick out which features are useful for improving performance.[6]

For supervised learning tasks, deep learning methods obviate feature engineering, by translating the data into compact intermediate representations akin to principal components, and derive layered structures that remove redundancy in representation.[3]

Deep learning algorithms can be applied to unsupervised learning tasks. This is an important benefit because unlabeled data are more abundant than labeled data. Examples of deep structures that can be trained in an unsupervised manner are neural history compressors[7] and deep belief networks.[6][8]

## Interpretations

Deep neural networks are generally interpreted in terms of the universal approximation theorem[9][10][11][12][13] or probabilistic inference.[3][4][6][5][8][14]

The universal approximation theorem concerns the capacity of feedforward neural networks with a single hidden layer of finite size to approximate continuous functions.[9][10][11][12][13]

In 1989, the first proof was published by Cybenko for sigmoid activation functions[10] and was generalised to feed-forward multi-layer architectures in 1991 by Hornik.[11]

The probabilistic interpretation[14] derives from the field of machine learning. It features inference,[3][4][6][5][8][14] as well as the optimization concepts of training and testing, related to fitting and generalization, respectively. More specifically, the probabilistic interpretation considers the activation nonlinearity as a cumulative distribution function.[14] The probabilistic interpretation led to the introduction of dropout as regularizer in neural networks.[15]

The probabilistic interpretation was introduced and popularized by Hinton, Bengio, LeCun and Schmidhuber.

## History

The first general, working learning algorithm for supervised, deep, feedforward, multilayer perceptrons was published by Ivakhnenko and Lapa in 1965.[16] A 1971 paper[17] described a deep network with 8 layers trained by the group method of data handling algorithm.[citation needed] These ideas were implemented in a computer identification system by the World School Council London called "Alpha", which demonstrated the learning process.

Other deep learning working architectures, specifically those built from artificial neural networks (ANN), began with the Neocognitron introduced by Fukushima in 1980.[18] ANNs date back even further. The challenge was how to train networks with multiple layers. In 1989, LeCun et al. applied the standard backpropagation algorithm, which had been around as the reverse mode of automatic differentiation since 1970,[19][20][21][22] to a deep neural network with the purpose of recognizing handwritten ZIP codes on mail. While the algorithm worked the training time was an impractical 3 days.[23]

In 1993, Schmidhuber's neural history compressor[24] implemented as an unsupervised stack of recurrent neural networks (RNNs) solved a "Very Deep Learning" task[5] that required more than 1,000 layers in an RNN unfolded in time.[25]

In 1994, André C. P. L. F. de Carvalho, together with Fairhurst and Bisset, published experimental results of a multi-layer boolean neural network, also known as a weightless neural network, composed of a self-organising feature extraction neural network module followed by a classification neural network module, which were independently trained.[26]

In 1995, Frey demonstrated that it was possible to train a network containing six fully connected layers and several hundred hidden units using the wake-sleep algorithm, co-developed with Dayan and Hinton.[27] However, training took two days. Many factors contribute to the slow speed, including the vanishing gradient problem analyzed in 1991 by Hochreiter.[28][29]

By 1991 such systems were used for recognizing isolated 2-D hand-written digits, while recognizing 3-D objects was done by matching 2-D images with a handcrafted 3-D object model. Weng et al. suggested that a human brain does not use a monolithic 3-D object model and in 1992 they published Cresceptron,[30][31][32] a method for performing 3-D object recognition directly from cluttered scenes. Cresceptron is a cascade of layers similar to Neocognitron. But while Neocognitron required a human programmer to hand-merge features, Cresceptron automatically learned an open number of unsupervised features in each layer, where each feature is represented by a convolution kernel. Cresceptron segmented each learned object from a cluttered scene through back-analysis through the network. Max pooling, now often adopted by deep neural networks (e.g. ImageNet tests), was first used in Cresceptron to reduce the position resolution by a factor of (2x2) to 1 through the cascade for better generalization.

Simpler models that use task-specific handcrafted features such as Gabor filters and support vector machines (SVMs) were a popular choice in the 1990s and 2000s, because of ANNs' computational cost and a lack of understanding of how the brain wires its biological networks.

Both shallow and deep learning (e.g., recurrent nets) of ANNs have been explored for many years.[33][34][35] These methods never outperformed non-uniform internal-handcrafting Gaussian mixture model/Hidden Markov model (GMM-HMM) technology based on generative models of speech trained discriminatively.[36] Key difficulties have been methodologically analyzed, including gradient diminishing[28] and weak temporal correlation structure in neural predictive models.[37][38] Additional difficulties were the lack of big training data and weaker computing power.

Most speech recognition researchers moved away from neural nets to pursue generative modeling. An exception was at SRI International in the late 1990s. Funded by the US government's NSA and DARPA, SRI conducted research on deep neural networks in speech and speaker recognition. Heck's speaker recognition team achieved the first significant success with deep neural networks in speech processing as demonstrated in the 1998 National Institute of Standards and Technology Speaker Recognition evaluation and later published in the journal of Speech Communication.[39] While SRI experienced success with deep neural networks in speaker recognition, they were unsuccessful in demonstrating similar success in speech recognition. One decade later, Hinton and Deng collaborated with each other and then with colleagues across groups at University of Toronto, Microsoft, Google and IBM, igniting a renaissance of deep feedforward neural networks in speech recognition.[40][41][42][43]

Many aspects of speech recognition were taken over by a deep learning method called Long short-term memory (LSTM), a recurrent neural network published by Hochreiter and Schmidhuber in 1997.[44] LSTM RNNs avoid the vanishing gradient problem and can learn "Very Deep Learning" tasks[5] that require memories of events that happened thousands of discrete time steps ago, which is important for speech. In 2003, LSTM started to become competitive with traditional speech recognizers on certain tasks.[45] Later it was combined with connectionist temporal classification (CTC)[46] in stacks of LSTM RNNs.[47] In 2015, Google's speech recognition reportedly experienced a dramatic performance jump of 49% through CTC-trained LSTM, which they made available through Google Voice Search.[48]

The use of the expression "Deep Learning" in the context of ANNs was introduced by Aizenberg and colleagues in 2000.[49] In 2006, Hinton and Salakhutdinov showed how a many-layered feedforward neural network could be effectively pre-trained one layer at a time, treating each layer in turn as an unsupervised restricted Boltzmann machine, then fine-tuning it using supervised backpropagation.[50] In 1992, Schmidhuber implemented a similar idea for the more general case of unsupervised deep hierarchies of recurrent neural networks and showed its benefits for accelerating supervised learning.[24][51]

Deep learning is part of state-of-the-art systems in various disciplines, particularly computer vision and automatic speech recognition (ASR). Results on commonly used evaluation sets such as TIMIT (ASR) and MNIST (image classification), as well as a range of large-vocabulary speech recognition tasks have steadily improved.[40][52][53] Convolutional neural networks have been superseded for ASR by CTC[46] for LSTM.[44][48][54][55][56][57][58] but are more successful in computer vision.

The impact of deep learning in industry began in the early 2000s, when CNNs already processed an estimated 10% to 20% of all the checks written in the US.[59] Industrial applications of deep learning to large-scale speech recognition started around 2010.

In late 2009, Li Deng invited Geoff Hinton to work with him and colleagues at Microsoft to apply deep learning to speech recognition. They co-organized the 2009 NIPS Workshop on Deep Learning for Speech Recognition [60] . The workshop was motivated by the limitations of deep generative models of speech, and the possibility that the big-compute, big-data era warranted a serious try of deep neural nets (DNN). It was believed that pre-training DNNs using generative models of deep belief nets (DBN) would overcome the main difficulties of neural nets.[42] However, Deng et al. at Microsoft soon after discovered that replacing pre-training with large amounts of training data for straightforward backpropagation when using DNNs with large, context-dependent output layers produced error rates dramatically lower than then-state-of-the-art Gaussian mixture model (GMM)/Hidden Markov Model (HMM) and also than more advanced generative model-based speech recognition systems. This finding was verified by other research groups subsequently.[40][61] Further, the nature of recognition errors produced by the two types of systems was found to be characteristically different,[41][60] offering technical insights into how to integrate deep learning into the existing highly efficient, run-time speech decoding system deployed by all major speech recognition players.[3][62][63]

Advances in hardware enabled the renewed interest in deep learning. In 2009, Nvidia was involved in what was called the “big bang” of deep learning, “as deep-learning neural networks were combined with Nvidia graphics processing units (GPUs).”[64] That year, Google Brain used Nvidia GPUs to create Deep Neural Networks capable of machine learning. While there Ng determined that GPUs could increase the speed of deep-learning systems by about 100 times.[65] In particular, powerful graphics processing units (GPUs) are well-suited for the matrix/vector math involved in machine learning.[66][67] GPUs speed up training algorithms by orders of magnitude, reducing running times from weeks to days.[68][69] Specialized hardware and algorithm optimizations can be used for efficient D processing.[70]

## Artificial neural networks

Artificial neural networks (ANNs) or connectionist systems are computing systems inspired by the biological neural networks that constitute animal brains. Such systems learn (progressively improve performance) to do tasks by considering examples, generally without task-specific programming. For example, in image recognition, they might learn to identify images that contain cats by analyzing example images that have been manually labeled as "cat" or "no cat" and using the analytic results to identify cats in other images. They have found most use in applications difficult to express in a traditional computer algorithm using rule-based programming.

An ANN is based on a collection of connected units called artificial neurons, (analogous to axons in a biological brain). Each connection (synapse) between neurons can transmit a signal to another neuron. The receiving (postsynaptic) neuron can process the signal(s) and then signal downstream neurons connected to it. Neurons may have state, generally represented by real numbers, typically between 0 and 1. Neurons and synapses may also have a weight that varies as learning proceeds, which can increase or decrease the strength of the signal that it sends downstream. Further, they may have a threshold such that only if the aggregate signal is below (or above) that level is the downstream signal sent.

Typically, neurons are organized in layers. Different layers may perform different kinds of transformations on their inputs. Signals travel from the first (input), to the last (output) layer, possibly after traversing the layers multiple times.

The original goal of the neural network approach was to solve problems in the same way that a human brain would. Over time, attention focused on matching specific mental abilities, leading to deviations from biology such as backpropagation, or passing information in the reverse direction and adjusting the network to reflect that information.

Neural networks have been used on a variety of tasks, including computer visionspeech recognitionmachine translationsocial network filtering, playing board and video games, medical diagnosis and in many other domains.

As of 2017, neural networks typically have a few thousand to a few million units and millions of connections. Their computing power is similar to a worm brain,[citation needed] several orders of magnitude simpler than a human brain. Despite this, they can perform functions (e.g., playing chess) that are far beyond a worm's capacity.

In 2007, LSTM[44] trained by CTC[46] started to get much-improved results.[47] This method is now widely used, for example, in Google's greatly improved speech recognition for all smartphone users.[48]

## Deep neural networks

A deep neural network (DNN) is an ANN with multiple hidden layers between the input and output layers.[4][5] Similar to shallow ANNs, DNNs can model complex non-linear relationships. DNN architectures generate compositional models where the object is expressed as a layered composition of primitives.[71] The extra layers enable composition of features from lower layers, potentially modeling complex data with fewer units than a similarly performing shallow network.[4]

Deep architectures include many variants of a few basic approaches. Each architecture has found success in specific domains. It is not always possible to compare the performance of multiple architectures, unless they have been evaluated on the same data sets.

DNNs are typically feedforward networks in which data flows from the input layer to the output layer without looping back. However, recurrent neural networks, in which data can flow in any direction, area also used, especially long short-term memory,[44][72] for applications such as language modeling.[73][74][75][76][77] Convolutional deep neural networks (CNNs) are used in computer vision.[78] CNNs also have been applied to acoustic modeling for automatic speech recognition (ASR).[58]

### Challenges

As with ANNs, many issues can arise with naively trained DNNs. Two common issues are overfitting and computation time.

DNNs are prone to overfitting because of the added layers of abstraction, which allow them to model rare dependencies in the training data. Regularization methods such as Ivakhnenko's unit pruning[17] or weight decay (${\displaystyle \ell _{2}}$-regularization) or sparsity (${\displaystyle \ell _{1}}$-regularization) can be applied during training to combat overfitting.[79] Alternatively dropout regularization randomly omits units from the hidden layers during training. This helps to exclude rare dependencies.[80]

DNNs must consider many training parameters, such as the size (number of layers and number of units per layer), the learning rate and initial weights. Sweeping through the parameter space for optimal parameters may not be feasible due to the cost in time and computational resources. Various tricks such as batching (computing the gradient on several training examples at once rather than individual examples)[81] speed up computation. The large processing throughput of GPUs has produced significant speedups in training, because the matrix and vector computations required are well-suited for GPUs.[5]

## Applications

### Automatic speech recognition

Speech recognition was revolutionised by deep learning, especially by long short-term memory RNNs.[44] LSTM RNNs circumvent the vanishing gradient problem and can learn "Very Deep Learning" tasks[5] that involve multi-second intervals containing speech events separated by thousands of discrete time steps, where one time step corresponds to about 10 ms. In 2003, LSTM with forget gates[72] became competitive with traditional speech recognizers on certain tasks.[45] In 2007, LSTM trained by CTC[46] achieved excellent results on tasks such as discriminative keyword spotting.[47] In 2015, Google's speech recognition almost doubled its performance through CTC-trained LSTM.[48]

The initial success in speech recognition, was based on small-scale recognition tasks based on the popular TIMIT data set (a common data set used for evaluations). The set contains 630 speakers from eight major dialects of American English, where each speaker reads 10 sentences.[82] Its small size allows many configurations to be tried. More importantly, the TIMIT task concerns phone-sequence recognition, which, unlike word-sequence recognition, allows very weak "language models".[clarification needed] This allows the weaknesses in acoustic modeling aspects of speech recognition to be more easily analyzed. Analysis on TIMIT by Li and collaborators around 2009-2010, contrasting the GMM (and other generative speech models) vs. DNN models, stimulated early industrial investment in deep learning for speech recognition,[41][60] eventually leading to pervasive and dominant use in that industry. That analysis was done with comparable performance (less than 1.5% in error rate) between discriminative DNNs and generative models. The error rates listed below, including these early results and measured as percent phone error rates (PER), have been summarized over the past 20 years:

Method PER (%)
Randomly Initialized RNN 26.1
Bayesian Triphone GMM-HMM 25.6
Hidden Trajectory (Generative) Model 24.8
Monophone Randomly Initialized DNN 23.4
Monophone DBN-DNN 22.4
Triphone GMM-HMM with BMMI Training 21.7
Monophone DBN-DNN on fbank 20.7
Convolutional DNN[83] 20.0
Convolutional DNN w. Heterogeneous Pooling 18.7
Ensemble DNN/CNN/RNN[84] 18.2
Bidirectional LSTM 17.9

In 2010, researchers extended deep learning from TIMIT to large vocabulary speech recognition, by adopting large output layers of the DNN based on context-dependent HMM states constructed by decision trees.[85][86][87][62]

The principle of elevating "raw" features over hand-crafted optimization was first explored successfully in the architecture of deep autoencoder on the "raw" spectrogram or linear filter-bank features in the late 1990s,[39] showing its superiority over the Mel-Cepstral features that contain stages of fixed transformation from spectrograms. The raw features of speech, waveforms, later produced excellent larger-scale results.[88]

The debut of DNNs for speaker recognition in the late 1990s and speech recognition around 2009-2011 and of LSTM around 2003-2007, accelerated progress in eight major areas:[3][43][62]

• Scale-up/out and acclerated DNN training and decoding
• Sequence discriminative training
• Feature processing by deep models with solid understanding of the underlying mechanisms
• Adaptation of DNNs and related deep models
• Multi-task and transfer learning by DNNs and related deep models
• CNNs and how to design them to best exploit domain knowledge of speech
• RNN and its rich LSTM variants
• Other types of deep models including tensor-based models and integrated deep generative/discriminative models.

Large-scale automatic speech recognition is the first and most convincing successful case of deep learning. Between 2010 and 2014, the two major conferences on signal processing and speech recognition, IEEE-ICASSP and Interspeech both saw a large increase in the numbers of accepted papers in their respective annual conference papers on the topic of deep learning for speech recognition. All major commercial speech recognition systems (e.g., Microsoft Cortana, Xbox, Skype Translator, Amazon Alexa, Google Now, Apple Siri, Baidu and iFlyTek voice search, and a range of Nuance speech products, etc.) are based on deep learning.[3][89][90][91]

### Image recognition

A common evaluation set for image classification is the MNIST database data set. MNIST is composed of handwritten digits and includes 60,000 training examples and 10,000 test examples. As with TIMIT, its small size allows multiple configurations to be tested. A comprehensive list of results on this set is available.[92] The current best result on MNIST is an error rate of 0.23%, achieved by Ciresan et al. in 2012.[93]

According to LeCun,[59] in the early 2000s, CNNs processed an estimated 10% to 20% of all the checks written in the US.

Significant additional impacts in image or object recognition were felt from 2011–2012. Although CNNs trained by backpropagation had been around for decades,[23] and GPU implementations of NNs for years,[66] including CNNs,[67] fast implementations of CNNs with max-pooling on GPUs in the style of Ciresan and colleaguesg[94] were needed to progress on computer vision.[5] In 2011, this approach achieved for the first time superhuman performance in a visual pattern recognition contest.[95] Also in 2011, it won the ICDAR Chinese handwriting contest, and in May 2012, it won the ISBI image segmentation contest.[96] Until 2011, CNNs did not play a major role at computer vision conferences, but in June 2012, a paper by Ciresan et al. at the leading conference CVPR[93] showed how max-pooling CNNs on GPU can dramatically improve many vision benchmark records. In October 2012, a similar system by Krizhevsky and Hinton[2] won the large-scale ImageNet competition by a significant margin over shallow machine learning methods. In November 2012, Ciresan et al.'s system also won the ICPR contest on analysis of large medical images for cancer detection, and in the following year also the MICCAI Grand Challenge on the same topic.[97] In 2013 and 2014, the error rate on the ImageNet task using deep learning was further reduced, following a similar trend in large-scale speech recognition. The Wolfram Image Identification project publicized these improvements.[98]

Image classification was then extended to the more challenging task of generating descriptions (captions) for images, often as a combination of CNNs and LSTMs.[99][100][101][102] Deep learning-trained vehicles now interpret 360° camera views.[103] Another example is Facial Dysmorphology Novel Analysis (FDNA) used to analyze cases of human malformation connected to a large database of genetic syndromes.

### Natural language processing

Neural networks have been used for implementing language models since the early 2000s.[73][104] Recurrent neural networks, especially LSTM,[44] are most appropriate for sequential data such as language. LSTM helped to improve machine translation[74] and language modeling.[75][76]

Other key techniques in this field are negative sampling[105] and word embedding. Word embedding, such as word2vec, can be thought of as a representational layer in a deep learning architecture that transforms an atomic word into a positional representation of the word relative to other words in the dataset; the position is represented as a point in a vector space. Using word embedding as an RNN input layer allows the network to parse sentences and phrases using an effective compositional vector grammar. A compositional vector grammar can be thought of as probabilistic context free grammar (PCFG) implemented by an RNN.[106] Recursive auto-encoders built atop word embeddings can assess sentence similarity and detect paraphrasing.[106] Deep neural architectures have achieved state-of-the-art results in natural language processing tasks such as constituency parsing,[107] sentiment analysis,[108] information retrieval,[109][110] spoken language understanding,[111] machine translation,[74][112] contextual entity linking,[113] writing style recognition[114] and others.[115][excessive citations]

### Drug discovery and toxicology

A large percentage of candidate drugs fail to win regulatory approval. These failures are caused by insufficient efficacy (on-target effect), undesired interactions (off-target effects), or unanticipated toxic effects.[116][117] In 2012, a team led by Dahl won the "Merck Molecular Activity Challenge" using multi-task deep neural networks to predict the biomolecular target of one drug.[118][119] In 2014, Hochreiter's group used deep learning to detect off-target and toxic effects of environmental chemicals in nutrients, household products and drugs and won the "Tox21 Data Challenge" of NIH, FDA and NCATS.[120][121][122] Deep learning may outdo other virtual screening methods.[123][124] Researchers enhanced deep learning for drug discovery by combining data from a variety of sources.[125] In 2015, Atomwise introduced AtomNet, the first deep learning neural networks for structure-based rational drug design.[126] Subsequently, AtomNet was used to predict novel candidate biomolecules for several disease targets, most notably treatments for the Ebola virus[127] and multiple sclerosis.[128][129]

### Customer relationship management

Deep reinforcement learning demonstrated a use in direct marketing settings, illustrating suitability for CRM automation. A neural network was used to approximate the value of possible direct marketing actions over the customer state space, defined in terms of RFM variables. The estimated value function was shown to have a natural interpretation as customer lifetime value.[130]

### Recommendation systems

Recommendation systems have used deep learning to extract meaningful features for a latent factor model for content-based music recommendations.[131] Multiview deep learning has been applied for learning user preferences from multiple domains.[132] The model uses a hybrid collaborative and content-based approach and enhances recommendations in multiple tasks.

### Bioinformatics

An autoencoder ANN was used in bioinformatics, to predict gene ontology annotations and gene-function relationships.[133]

In medical informatics, deep learning was used to predict sleep quality based on data from wearables[134] and predictions of health complications from Electronic health record data.[135]

## Relation to human development

Deep learning is closely related to a class of theories of brain development (specifically, neocortical development) proposed by cognitive neuroscientists in the early 1990s.[136] An approachable summary of this work is Elman, et al.'s 1996 Rethinking Innateness[137] (see also: Shrager and Johnson;[138] Quartz and Sejnowski[139]). These developmental theories were instantiated in computational models, making them predecessors of purely computation-derived deep learning models. These developmental models share the property that various proposed learning dynamics in the brain (e.g., a wave of nerve growth factor) support the self-organization similar to the inter-related neural networks utilized in deep learning models. Such computational neural networks seem analogous to a view of the neocortex as a hierarchy of filters in which each layer captures some of the information in the operating environment, and then passes the remainder, as well as modified base signal, to other layers further up the hierarchy. This process yields a self-organizing stack of transducers, well-tuned to their operating environment. A 1995 description stated, "...the infant's brain seems to organize itself under the influence of waves of so-called trophic-factors ... different regions of the brain become connected sequentially, with one layer of tissue maturing before another and so on until the whole brain is mature."[140]

## Commercial activity

Many organizations have become interested in deep learning for particular applications. In 2013, Facebook hired Yann LeCun to head its new artificial intelligence (AI) lab. The AI lab will help perform tasks such as automatically tagging uploaded pictures with the names of the people in them.[141] In 2014, Facebook hired Vladimir Vapnik, a main developer of the Vapnik–Chervonenkis theory of statistical learning, and co-inventor of the support vector machine method.[142]

In 2014, Google bought DeepMind Technologies, a British start-up that developed a system capable of learning how to play Atari video games using only pixels as data input. In 2015 they demonstrated their AlphaGo system which achieved one of the long-standing "grand challenges" of AI by learning the game of Go well enough to beat a professional Go player.[143][144][145]

In 2015, Blippar demonstrated a mobile augmented reality application that uses deep learning to recognize objects in real time.[146]

## Criticism and comment

Deep learning has attracted both criticism and comment, in some cases from outside the field of computer science.

A main criticism concerns the lack of theory surrounding the methods.[citation needed] Learning in the most common deep architectures is implemented using well-understood gradient descent. However, the theory surrounding other algorithms, such as contrastive divergence is less clear.[citation needed] (e.g., Does it converge? If so, how fast? What is it approximating?) Deep learning methods are often looked at as a black box, with most confirmations done empirically, rather than theoretically.[citation needed]

Others point out that deep learning should be looked at as a step towards realizing strong AI, not as an all-encompassing solution. Despite the power of deep learning methods, they still lack much of the functionality needed for realizing this goal entirely. Research psychologist Gary Marcus noted:

"Realistically, deep learning is only part of the larger challenge of building intelligent machines. Such techniques lack ways of representing causal relationships (...) have no obvious ways of performing logical inferences, and they are also still a long way from integrating abstract knowledge, such as information about what objects are, what they are for, and how they are typically used. The most powerful A.I. systems, like Watson (...) use techniques like deep learning as just one element in a very complicated ensemble of techniques, ranging from the statistical technique of Bayesian inference to deductive reasoning."[147]

Alternatively, one author speculated that it might be possible to train a machine vision stack to perform the sophisticated task of discriminating between "old master" and amateur figure drawings; while another hypothesized that such a sensitivity might represent the rudiments of a non-trivial machine empathy.[148] Another author proposed that this would be in line with anthropology, which identifies a concern with aesthetics as a key element of behavioral modernity.[149]

In further reference to the idea that artistic sensitivity might inhere within relatively low levels of the cognitive hierarchy, a published series of graphic representations of the internal states of deep (20-30 layers) neural networks attempting to discern within essentially random data the images on which they were trained[150] demonstrate a visual appeal: the original research notice received well over 1,000 comments, and was for a time the most frequently accessed article on The Guardian's[151] web site.

Some deep learning architectures display problematic behaviors,[152] such as confidently classifying unrecognizable images as belonging to a familiar category of ordinary images[153] and misclassifying minuscule perturbations of correctly classified images.[154] Goertzel hypothesized that these behaviors are due to limitations in their internal representations and that these limitations would inhibit integration into heterogeneous multi-component AGI architectures.[152] These issues may possibly be addresed by deep learning architectures that internally form states homologous to image-grammar[155] decompositions of observed entities and events.[152] Learning a grammar (visual or linguistic) from training data would be equivalent to restricting the system to commonsense reasoning that operates on concepts in terms of grammatical production rules and is a basic goal of both human language acquisition[156] and AI.[157]

## Software libraries

• Deeplearning4j — An open-source deep-learning library written for Java/C++ with LSTMs and convolutional networks. It provides parallelization with Spark on CPUs and GPUs.
• Gensim — A toolkit for natural language processing implemented in the Python programming language.
• Keras — An open-source deep learning framework for the Python programming language.
• Microsoft CNTK (Computational Network Toolkit) — Microsoft's open-source deep-learning toolkit for Windows and Linux. It provides parallelization with CPUs and GPUs across multiple servers.
• MXNet — An open source deep learning framework that allows you to define, train, and deploy deep neural networks.
• OpenNN — An open source C++ library which implements deep neural networks and provides parallelization with CPUs.
• PaddlePaddle — An open source C++ /CUDA library with Python API for scalable deep learning platform with CPUs and GPUs, originally developed by Baidu.
• TensorFlow — Google's open source machine learning library in C++ and Python with APIs for both. It provides parallelization with CPUs and GPUs.
• Theano — An open source machine learning library for Python supported by the University of Montreal and Yoshua Bengio's team.
• Torch — An open source software library for machine learning based on the Lua programming language and used by Facebook.
• Caffe - Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by the Berkeley Vision and Learning Center (BVLC) and by community contributors.
• DIANNE - A modular open-source deep learning framework in Java / OSGi developed at Ghent University, Belgium. It provides parallelization with CPUs and GPUs across multiple servers.

## References

1. ^ Olshausen, B. A. (1996). "Emergence of simple-cell receptive field properties by learning a sparse code for natural images". Nature. 381 (6583): 607–609. PMID 8637596. doi:10.1038/381607a0.
2. ^ a b Krizhevsky, Alex; Sutskever, Ilya; Hinton, Geoffry (2012). "ImageNet Classification with Deep Convolutional Neural Networks" (PDF). NIPS 2012: Neural Information Processing Systems, Lake Tahoe, Nevada.
3. Deng, L.; Yu, D. (2014). "Deep Learning: Methods and Applications" (PDF). Foundations and Trends in Signal Processing. 7 (3–4): 1–199. doi:10.1561/2000000039.
4. Bengio, Yoshua (2009). "Learning Deep Architectures for AI" (PDF). Foundations and Trends in Machine Learning. 2 (1): 1–127. doi:10.1561/2200000006.
5. Schmidhuber, J. (2015). "Deep Learning in Neural Networks: An Overview". Neural Networks. 61: 85–117. arXiv:. doi:10.1016/j.neunet.2014.09.003.
6. Bengio, Y.; Courville, A.; Vincent, P. (2013). "Representation Learning: A Review and New Perspectives". IEEE Transactions on Pattern Analysis and Machine Intelligence. 35 (8): 1798–1828. arXiv:. doi:10.1109/tpami.2013.50.
7. ^ Schmidhuber, Jürgen (1992-03-01). "Learning Complex, Extended Sequences Using the Principle of History Compression". Neural Computation. 4 (2): 234-242. ISSN 0899-7667. doi:10.1162/neco.1992.4.2.234.
8. ^ a b c Hinton, G.E. "Deep belief networks". Scholarpedia. 4 (5): 5947. doi:10.4249/scholarpedia.5947.
9. ^ a b Balázs Csanád Csáji (2001). Approximation with Artificial Neural Networks; Faculty of Sciences; Eötvös Loránd University, Hungary
10. ^ a b c Cybenko (1989). "Approximations by superpositions of sigmoidal functions" (PDF). Mathematics of Control, Signals, and Systems. 2 (4): 303–314. doi:10.1007/bf02551274.
11. ^ a b c Hornik, Kurt (1991). "Approximation Capabilities of Multilayer Feedforward Networks". Neural Networks. 4 (2): 251–257. doi:10.1016/0893-6080(91)90009-t.
12. ^ a b Haykin, Simon S. (1999). Neural Networks: A Comprehensive Foundation. Prentice Hall. ISBN 978-0-13-273350-2.
13. ^ a b Hassoun, Mohamad H. (1995). Fundamentals of Artificial Neural Networks. MIT Press. p. 48. ISBN 978-0-262-08239-6.
14. ^ a b c d Murphy, Kevin P. (24 August 2012). Machine Learning: A Probabilistic Perspective. MIT Press. ISBN 978-0-262-01802-9.
15. ^ Hinton, G. E.; Srivastava, N.; Krizhevsky, A.; Sutskever, I.; Salakhutdinov, R.R. (2012). "Improving neural networks by preventing co-adaptation of feature detectors". arXiv: [math.LG].
16. ^ Ivakhnenko, A. G. (1973). Cybernetic Predicting Devices. CCM Information Corporation.
17. ^ a b Ivakhnenko, Alexey (1971). "Polynomial theory of complex systems". IEEE Transactions on Systems, Man and Cybernetics (4): 364–378. doi:10.1109/TSMC.1971.4308320.
18. ^ Fukushima, K. (1980). "Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position". Biol. Cybern. 36: 193–202. PMID 7370364. doi:10.1007/bf00344251.
19. ^ Seppo Linnainmaa (1970). The representation of the cumulative rounding error of an algorithm as a Taylor expansion of the local rounding errors. Master's Thesis (in Finnish), Univ. Helsinki, 6-7.
20. ^ Griewank, Andreas (2012). "Who Invented the Reverse Mode of Differentiation?" (PDF). Documenta Matematica, Extra Volume ISMP: 389–400.
21. ^ Werbos, P. (1974). "Beyond Regression: New Tools for Prediction and Analysis in the Behavioral Sciences". Harvard University. Retrieved 12 June 2017.
22. ^ Werbos, Paul (1982). "Applications of advances in nonlinear sensitivity analysis". System modeling and optimization (PDF). Springer. pp. 762–770.
23. ^ a b LeCun et al., "Backpropagation Applied to Handwritten Zip Code Recognition," Neural Computation, 1, pp. 541–551, 1989.
24. ^ a b J. Schmidhuber., "Learning complex, extended sequences using the principle of history compression," Neural Computation, 4, pp. 234–242, 1992.
25. ^ Jürgen Schmidhuber (1993). Habilitation thesis, TUM, 1993. Page 150 ff demonstrates credit assignment across the equivalent of 1,200 layers in an unfolded RNN. Online
26. ^ de Carvalho, Andre C. L. F.; Fairhurst, Mike C.; Bisset, David (1994-08-08). "An integrated Boolean neural network for pattern classification". Pattern Recognition Letters. 15 (8): 807–813. doi:10.1016/0167-8655(94)90009-4.
27. ^ Hinton, Geoffrey E.; Dayan, Peter; Frey, Brendan J.; Neal, Radford (1995-05-26). "The wake-sleep algorithm for unsupervised neural networks". Science. 268 (5214): 1158–1161. doi:10.1126/science.7761831.
28. ^ a b S. Hochreiter., "Untersuchungen zu dynamischen neuronalen Netzen," Diploma thesis. Institut f. Informatik, Technische Univ. Munich. Advisor: J. Schmidhuber, 1991.
29. ^ Hochreiter, S.; et al. (15 January 2001). "Gradient flow in recurrent nets: the difficulty of learning long-term dependencies". In Kolen, John F.; Kremer, Stefan C. A Field Guide to Dynamical Recurrent Networks. John Wiley & Sons. ISBN 978-0-7803-5369-5.
30. ^ J. Weng, N. Ahuja and T. S. Huang, "Cresceptron: a self-organizing neural network which grows adaptively," Proc. International Joint Conference on Neural Networks, Baltimore, Maryland, vol I, pp. 576-581, June, 1992.
31. ^ J. Weng, N. Ahuja and T. S. Huang, "Learning recognition and segmentation of 3-D objects from 2-D images," Proc. 4th International Conf. Computer Vision, Berlin, Germany, pp. 121-128, May, 1993.
32. ^ J. Weng, N. Ahuja and T. S. Huang, "Learning recognition and segmentation using the Cresceptron," International Journal of Computer Vision, vol. 25, no. 2, pp. 105-139, Nov. 1997.
33. ^ Morgan, Nelson; Bourlard, Hervé; Renals, Steve; Cohen, Michael; Franco, Horacio (1993-08-01). "Hybrid neural network/hidden markov model systems for continuous speech recognition". International Journal of Pattern Recognition and Artificial Intelligence. 07 (04): 899–916. ISSN 0218-0014. doi:10.1142/s0218001493000455.
34. ^ Robinson, T. (1992). "A real-time recurrent error propagation network word recognition system". ICASSP.
35. ^ Waibel, A.; Hanazawa, T.; Hinton, G.; Shikano, K.; Lang, K. J. (March 1989). "Phoneme recognition using time-delay neural networks". IEEE Transactions on Acoustics, Speech, and Signal Processing. 37 (3): 328–339. ISSN 0096-3518. doi:10.1109/29.21701.
36. ^ Baker, J.; Deng, Li; Glass, Jim; Khudanpur, S.; Lee, C.-H.; Morgan, N.; O'Shaughnessy, D. (2009). "Research Developments and Directions in Speech Recognition and Understanding, Part 1". IEEE Signal Processing Magazine. 26 (3): 75–80. doi:10.1109/msp.2009.932166.
37. ^ Bengio, Y. (1991). "Artificial Neural Networks and their Application to Speech/Sequence Recognition". McGill University Ph.D. thesis.
38. ^ Deng, L.; Hassanein, K.; Elmasry, M. (1994). "Analysis of correlation structure for a neural predictive model with applications to speech recognition". Neural Networks. 7 (2): 331–339. doi:10.1016/0893-6080(94)90027-2.
39. ^ a b Heck, L.; Konig, Y.; Sonmez, M.; Weintraub, M. (2000). "Robustness to Telephone Handset Distortion in Speaker Recognition by Discriminative Feature Design". Speech Communication. 31 (2): 181–192. doi:10.1016/s0167-6393(99)00077-1.
40. ^ a b c Hinton, G.; Deng, L.; Yu, D.; Dahl, G.; Mohamed, A.; Jaitly, N.; Senior, A.; Vanhoucke, V.; Nguyen, P.; Sainath, T.; Kingsbury, B. (2012). "Deep Neural Networks for Acoustic Modeling in Speech Recognition --- The shared views of four research groups". IEEE Signal Processing Magazine. 29 (6): 82–97. doi:10.1109/msp.2012.2205597.
41. ^ a b c Deng, L.; Hinton, G.; Kingsbury, B. (2013). "New types of deep neural network learning for speech recognition and related applications: An overview (ICASSP)" (PDF).
42. ^ a b Keynote talk: Recent Developments in Deep Neural Networks. ICASSP, 2013 (by Geoff Hinton).
43. ^ a b Li, Deng (September 2014). "Keynote talk: 'Achievements and Challenges of Deep Learning - From Speech Analysis and Recognition To Language and Multimodal Processing'". Interspeech.
44. Hochreiter, Sepp; Schmidhuber, Jürgen (1997-11-01). "Long Short-Term Memory". Neural Computation. 9 (8): 1735–1780. ISSN 0899-7667. doi:10.1162/neco.1997.9.8.1735.
45. ^ a b Graves, Alex; Eck, Douglas; Beringer, Nicole; Schmidhuber, Jürgen (2003). "Biologically Plausible Speech Recognition with LSTM Neural Nets" (PDF). 1st Intl. Workshop on Biologically Inspired Approaches to Advanced Information Technology, Bio-ADIT 2004, Lausanne, Switzerland. pp. 175–184.
46. ^ a b c d Graves, Alex; Fernández, Santiago; Gomez, Faustino (2006). "Connectionist temporal classification: Labelling unsegmented sequence data with recurrent neural networks". In Proceedings of the International Conference on Machine Learning, ICML 2006: 369–376.
47. ^ a b c Santiago Fernandez, Alex Graves, and Jürgen Schmidhuber (2007). An application of recurrent neural networks to discriminative keyword spotting. Proceedings of ICANN (2), pp. 220–229.
48. ^ a b c d Sak, Haşim; Senior, Andrew; Rao, Kanishka; Beaufays, Françoise; Schalkwyk, Johan (September 2015). "Google voice search: faster and more accurate".
49. ^ N., Aizenberg, Naum; Joos., Vandewalle, (2000). Multi-Valued and Universal Binary Neurons : Theory, Learning and Applications. Springer US. ISBN 9781475731156. OCLC 851783812.
50. ^ Hinton, Geoffrey E. (2007-10-01). "Learning multiple layers of representation". Trends in Cognitive Sciences. 11 (10): 428–434. ISSN 1364-6613. PMID 17921042. doi:10.1016/j.tics.2007.09.004.
51. ^ J. Schmidhuber., "My First Deep Learning System of 1991 + Deep Learning Timeline 1962–2013." Online
52. ^ Deng, Li; Hinton, Geoffrey; Kingsbury, Brian (1 May 2013). "New types of deep neural network learning for speech recognition and related applications: An overview" – via research.microsoft.com.
53. ^ Deng, L.; Li, J.; Huang, J. T.; Yao, K.; Yu, D.; Seide, F.; Seltzer, M.; Zweig, G.; He, X. (May 2013). "Recent advances in deep learning for speech research at Microsoft". 2013 IEEE International Conference on Acoustics, Speech and Signal Processing: 8604–8608. doi:10.1109/icassp.2013.6639345.
54. ^ Sak, Hasim; Senior, Andrew; Beaufays, Francoise (2014). "Long Short-Term Memory recurrent neural network architectures for large scale acoustic modeling" (PDF).
55. ^ Xiangang Li, Xihong Wu (2015). Constructing Long Short-Term Memory based Deep Recurrent Neural Networks for Large Vocabulary Speech Recognition arXiv:1410.4281
56. ^ Zen, Heiga; Sak, Hasim (2015). "Unidirectional Long Short-Term Memory Recurrent Neural Network with Recurrent Output Layer for Low-Latency Speech Synthesis" (PDF). Google.com. ICASSP. pp. 4470–4474.
57. ^ Deng, L.; Abdel-Hamid, O.; Yu, D. (2013). "A deep convolutional neural network using heterogeneous pooling for trading acoustic invariance with phonetic confusion" (PDF). Google.com. ICASSP.
58. ^ a b Sainath, T. N.; Mohamed, A. r; Kingsbury, B.; Ramabhadran, B. (May 2013). "Deep convolutional neural networks for LVCSR". 2013 IEEE International Conference on Acoustics, Speech and Signal Processing: 8614–8618. doi:10.1109/icassp.2013.6639347.
59. ^ a b Yann LeCun (2016). Slides on Deep Learning Online
60. ^ a b c NIPS Workshop: Deep Learning for Speech Recognition and Related Applications, Whistler, BC, Canada, Dec. 2009 (Organizers: Li Deng, Geoff Hinton, D. Yu).
61. ^ D. Yu, L. Deng, G. Li, and F. Seide (2011). "Discriminative pretraining of deep neural networks," U.S. Patent Filing.
62. ^ a b c Yu, D.; Deng, L. (2014). "Automatic Speech Recognition: A Deep Learning Approach (Publisher: Springer)". ISBN 978-1-4471-5779-3.
63. ^
64. ^ "Nvidia CEO bets big on deep learning and VR". Venture Beat. April 5, 2016.
65. ^
66. ^ a b Oh, K.-S.; Jung, K. (2004). "GPU implementation of neural networks". Pattern Recognition. 37 (6): 1311–1314. doi:10.1016/j.patcog.2004.01.013.
67. ^ a b Chellapilla, K., Puri, S., and Simard, P. (2006). High performance convolutional neural networks for document processing. International Workshop on Frontiers in Handwriting Recognition.
68. ^ Cireşan, Dan Claudiu; Meier, Ueli; Gambardella, Luca Maria; Schmidhuber, Jürgen (2010-09-21). "Deep, Big, Simple Neural Nets for Handwritten Digit Recognition". Neural Computation. 22 (12): 3207–3220. ISSN 0899-7667. doi:10.1162/neco_a_00052.
69. ^ Raina, Rajat; Madhavan, Anand; Ng, Andrew Y. (2009). "Large-scale Deep Unsupervised Learning Using Graphics Processors". Proceedings of the 26th Annual International Conference on Machine Learning. ICML '09. New York, NY, USA: ACM: 873–880. ISBN 9781605585161. doi:10.1145/1553374.1553486.
70. ^ Sze, Vivienne; Chen, Yu-Hsin; Yang, Tien-Ju; Emer, Joel (2017). "Efficient Processing of Deep Neural Networks: A Tutorial and Survey". arXiv:.
71. ^ Szegedy, Christian; Toshev, Alexander; Erhan, Dumitru (2013). "Deep neural networks for object detection". Advances in Neural Information Processing Systems.
72. ^ a b "Learning Precise Timing with LSTM Recurrent Networks (PDF Download Available)". ResearchGate. Retrieved 2017-06-13.
73. ^ a b Gers, Felix A.; Schmidhuber, Jürgen (2001). "LSTM Recurrent Networks Learn Simple Context Free and Context Sensitive Languages". IEEE TNN. 12 (6): 1333–1340. doi:10.1109/72.963769.
74. ^ a b c Sutskever, L.; Vinyals, O.; Le, Q. (2014). "Sequence to Sequence Learning with Neural Networks" (PDF). Proc. NIPS.
75. ^ a b Rafal Jozefowicz, Oriol Vinyals, Mike Schuster, Noam Shazeer, Yonghui Wu (2016). Exploring the Limits of Language Modeling. arXiv
76. ^ a b Dan Gillick, Cliff Brunk, Oriol Vinyals, Amarnag Subramanya (2015). Multilingual Language Processing From Bytes. arXiv
77. ^ Mikolov, T.; et al. (2010). "Recurrent neural network based language model" (PDF). Interspeech.
78. ^ LeCun, Y.; et al. "Gradient-based learning applied to document recognition". Proceedings of the IEEE. 86 (11): 2278–2324. doi:10.1109/5.726791.
79. ^ Bengio, Y.; Boulanger-Lewandowski, N.; Pascanu, R. (May 2013). "Advances in optimizing recurrent networks". 2013 IEEE International Conference on Acoustics, Speech and Signal Processing: 8624–8628. doi:10.1109/icassp.2013.6639349.
80. ^ Dahl, G.; et al. (2013). "Improving DNNs for LVCSR using rectified linear units and dropout" (PDF). ICASSP.
81. ^ Hinton, G. E. (2010). "A Practical Guide to Training Restricted Boltzmann Machines". Tech. Rep. UTML TR 2010-003,.
82. ^ TIMIT Acoustic-Phonetic Continuous Speech Corpus Linguistic Data Consortium, Philadelphia.
83. ^ Abdel-Hamid, O.; et al. (2014). "Convolutional Neural Networks for Speech Recognition". IEEE/ACM Transactions on Audio, Speech, and Language Processing. 22 (10): 1533–1545. doi:10.1109/taslp.2014.2339736.
84. ^ Deng, L.; Platt, J. (2014). "Ensemble Deep Learning for Speech Recognition" (PDF). Proc. Interspeech.
85. ^ Yu, D.; Deng, L. (2010). "Roles of Pre-Training and Fine-Tuning in Context-Dependent DBN-HMMs for Real-World Speech Recognition". NIPS Workshop on Deep Learning and Unsupervised Feature Learning.
86. ^ Seide, F.; Li, G.; Yu, D. (2011). "Conversational speech transcription using context-dependent deep neural networks". Interspeech.
87. ^ Deng, Li; Li, Jinyu; Huang, Jui-Ting; Yao, Kaisheng; Yu, Dong; Seide, Frank; Seltzer, Mike; Zweig, Geoff; He, Xiaodong (2013-05-01). "Recent Advances in Deep Learning for Speech Research at Microsoft". Microsoft Research.
88. ^ "Acoustic Modeling with Deep Neural Networks Using Raw Time Signal for LVCSR (PDF Download Available)". ResearchGate. Retrieved 2017-06-14.
89. ^ "How Skype Used AI to Build Its Amazing New Language Translator | WIRED". www.wired.com. Retrieved 2017-06-14.
90. ^ Hannun et al. (2014) "Deep Speech: Scaling up end-to-end speech recognition", arXiv:1412.5567.
91. ^
92. ^
93. ^ a b Ciresan, Dan; Meier, U.; Schmidhuber, J. (June 2012). "Multi-column deep neural networks for image classification". 2012 IEEE Conference on Computer Vision and Pattern Recognition: 3642–3649. doi:10.1109/cvpr.2012.6248110.
94. ^ Ciresan, D. C.; Meier, U.; Masci, J.; Gambardella, L. M.; Schmidhuber, J. (2011). "Flexible, High Performance Convolutional Neural Networks for Image Classification" (PDF). International Joint Conference on Artificial Intelligence. doi:10.5591/978-1-57735-516-8/ijcai11-210.
95. ^ Cireşan, Dan; Meier, Ueli; Masci, Jonathan; Schmidhuber, Jürgen (August 2012). "Multi-column deep neural network for traffic sign classification". Neural Networks. Selected Papers from IJCNN 2011. 32: 333–338. doi:10.1016/j.neunet.2012.02.023.
96. ^ Ciresan, Dan; Giusti, Alessandro; Gambardella, Luca M.; Schmidhuber, Juergen (2012). Pereira, F.; Burges, C. J. C.; Bottou, L.; Weinberger, K. Q., eds. Advances in Neural Information Processing Systems 25 (PDF). Curran Associates, Inc. pp. 2843–2851.
97. ^ Ciresan, D.; Giusti, A.; Gambardella, L.M.; Schmidhuber, J. (2013). "Mitosis Detection in Breast Cancer Histology Images using Deep Neural Networks" (PDF). Proceedings MICCAI,.
98. ^ "The Wolfram Language Image Identification Project". www.imageidentify.com. Retrieved 2017-03-22.
99. ^ Vinyals et al. (2014)."Show and Tell: A Neural Image Caption Generator," arXiv:1411.4555.
100. ^ Fang et al. (2014)."From Captions to Visual Concepts and Back," arXiv:1411.4952.
101. ^ Kiros et al. (2014). "Unifying Visual-Semantic Embeddings with Multimodal Neural Language Models," arXiv:1411.2539.
102. ^ Zhong, Sheng-hua; Liu, Yan; Liu, Yang (2011). "Bilinear Deep Learning for Image Classification". Proceedings of the 19th ACM International Conference on Multimedia. MM '11. New York, NY, USA: ACM: 343–352. ISBN 9781450306164. doi:10.1145/2072298.2072344.
103. ^ Nvidia Demos a Car Computer Trained with "Deep Learning" (2015-01-06), David Talbot, MIT Technology Review
104. ^ Bengio, Yoshua; Ducharme, Réjean; Vincent, Pascal; Janvin, Christian (March 2003). "A Neural Probabilistic Language Model". J. Mach. Learn. Res. 3: 1137–1155. ISSN 1532-4435.
105. ^ Goldberg, Yoav; Levy, Omar. "word2vec Explained: Deriving Mikolov et al.’s Negative-Sampling Word-Embedding Method". arXiv:.
106. ^ a b Socher, Richard; Manning, Christopher. "Deep Learning for NLP" (PDF). Retrieved 26 October 2014.
107. ^ Socher, Richard; Bauer, John; Manning, Christopher; Ng, Andrew (2013). "Parsing With Compositional Vector Grammars" (PDF). Proceedings of the ACL 2013 conference.
108. ^ Socher, Richard (2013). "Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank" (PDF). EMNLP 2013.
109. ^ Shen, Yelong; He, Xiaodong; Gao, Jianfeng; Deng, Li; Mesnil, Gregoire (2014-11-01). "A Latent Semantic Model with Convolutional-Pooling Structure for Information Retrieval". Microsoft Research.
110. ^ Huang, Po-Sen; He, Xiaodong; Gao, Jianfeng; Deng, Li; Acero, Alex; Heck, Larry (2013-10-01). "Learning Deep Structured Semantic Models for Web Search using Clickthrough Data". Microsoft Research.
111. ^ Mesnil, G.; Dauphin, Y.; Yao, K.; Bengio, Y.; Deng, L.; Hakkani-Tur, D.; He, X.; Heck, L.; Tur, G.; Yu, D.; Zweig, G. (2015). "Using recurrent neural networks for slot filling in spoken language understanding". IEEE Transactions on Audio, Speech, and Language Processing. 23 (3): 530–539. doi:10.1109/taslp.2014.2383614.
112. ^ Gao, Jianfeng; He, Xiaodong; Yih, Scott Wen-tau; Deng, Li (2014-06-01). "Learning Continuous Phrase Representations for Translation Modeling". Microsoft Research.
113. ^ Gao, Jianfeng; He, Xiaodong; Yih, Scott Wen-tau; Deng, Li (2014-06-01). "Learning Continuous Phrase Representations for Translation Modeling". Microsoft Research.
114. ^ Brocardo ML, Traore I, Woungang I, Obaidat MS. "Authorship verification using deep belief network systems". Int J Commun Syst. 2017. doi:10.1002/dac.3259
115. ^ "Deep Learning for Natural Language Processing: Theory and Practice (CIKM2014 Tutorial) - Microsoft Research". Microsoft Research. Retrieved 2017-06-14.
116. ^ Arrowsmith, J; Miller, P (2013). "Trial watch: Phase II and phase III attrition rates 2011-2012". Nature Reviews Drug Discovery. 12 (8): 569. PMID 23903212. doi:10.1038/nrd4090.
117. ^ Verbist, B; Klambauer, G; Vervoort, L; Talloen, W; The Qstar, Consortium; Shkedy, Z; Thas, O; Bender, A; Göhlmann, H. W.; Hochreiter, S (2015). "Using transcriptomics to guide lead optimization in drug discovery projects: Lessons learned from the QSTAR project". Drug Discovery Today. 20: 505–513. PMID 25582842. doi:10.1016/j.drudis.2014.12.014.
118. ^
119. ^ "Multi-task Neural Networks for QSAR Predictions | Data Science Association". www.datascienceassn.org. Retrieved 2017-06-14.
120. ^ "Toxicology in the 21st century Data Challenge]
121. ^ NCATS Announces Tox21 Data Challenge Winners
122. ^ http://www.ncats.nih.gov/news-and-events/features/tox21-challenge-winners.html
123. ^ Unterthiner, T.; Mayr, A.; Klambauer, G.; Steijaert, M.; Ceulemans, H.; Wegner, J. K.; & Hochreiter, S. (2014) "Deep Learning as an Opportunity in Virtual Screening". Workshop on Deep Learning and Representation Learning (NIPS2014).
124. ^ Unterthiner, T.; Mayr, A.; Klambauer, G.; & Hochreiter, S. (2015) "Toxicity Prediction using Deep Learning". ArXiv, 2015.
126. ^ Wallach, Izhar; Dzamba, Michael; Heifets, Abraham (2015-10-09). "AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery". arXiv:.
127. ^ "Toronto startup has a faster way to discover effective medicines". The Globe and Mail. Retrieved 2015-11-09.
128. ^ "Startup Harnesses Supercomputers to Seek Cures". KQED Future of You. Retrieved 2015-11-09.
129. ^
130. ^ Tkachenko, Yegor (April 8, 2015). "Autonomous CRM Control via CLV Approximation with Deep Reinforcement Learning in Discrete and Continuous Action Space".
131. ^ van den Oord, Aaron; Dieleman, Sander; Schrauwen, Benjamin (2013). Burges, C. J. C.; Bottou, L.; Welling, M.; Ghahramani, Z.; Weinberger, K. Q., eds. Advances in Neural Information Processing Systems 26 (PDF). Curran Associates, Inc. pp. 2643–2651.
132. ^ Elkahky, Ali Mamdouh; Song, Yang; He, Xiaodong (2015-05-01). "A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems". Microsoft Research.
133. ^ Chicco, Davide; Sadowski, Peter; Baldi, Pierre (1 January 2014). "Deep Autoencoder Neural Networks for Gene Ontology Annotation Predictions". Proceedings of the 5th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics - BCB '14. ACM: 533–540. doi:10.1145/2649387.2649442 – via ACM Digital Library.
134. ^ Sathyanarayana, Aarti (2016-01-01). "Sleep Quality Prediction From Wearable Data Using Deep Learning". JMIR mHealth and uHealth. 4 (4): e125. doi:10.2196/mhealth.6562.
135. ^ Choi, Edward; Schuetz, Andy; Stewart, Walter F.; Sun, Jimeng (2016-08-13). "Using recurrent neural network models for early detection of heart failure onset". Journal of the American Medical Informatics Association: ocw112. ISSN 1067-5027. PMID 27521897. doi:10.1093/jamia/ocw112.
136. ^ Utgoff, P. E.; Stracuzzi, D. J. (2002). "Many-layered learning". Neural Computation. 14: 2497–2529. doi:10.1162/08997660260293319.
137. ^ Elman, Jeffrey L. (1998). Rethinking Innateness: A Connectionist Perspective on Development. MIT Press. ISBN 978-0-262-55030-7.
138. ^ Shrager, J.; Johnson, MH (1996). "Dynamic plasticity influences the emergence of function in a simple cortical array". Neural Networks. 9 (7): 1119–1129. doi:10.1016/0893-6080(96)00033-0.
139. ^ Quartz, SR; Sejnowski, TJ (1997). "The neural basis of cognitive development: A constructivist manifesto". Behavioral and Brain Sciences. 20 (4): 537–556. doi:10.1017/s0140525x97001581.
140. ^ S. Blakeslee., "In brain's early growth, timetable may be critical," The New York Times, Science Section, pp. B5–B6, 1995.
141. ^ Metz, C. (12 December 2013). "Facebook's 'Deep Learning' Guru Reveals the Future of AI". Wired.
143. ^ "Google AI algorithm masters ancient game of Go". Nature News & Comment. Retrieved 2016-01-30.
144. ^ Silver, David; Huang, Aja; Maddison, Chris J.; Guez, Arthur; Sifre, Laurent; van den Driessche, George; Schrittwieser, Julian; Antonoglou, Ioannis; Panneershelvam, Veda (2016-01-28). "Mastering the game of Go with deep neural networks and tree search". Nature. 529 (7587): 484–489. ISSN 0028-0836. PMID 26819042. doi:10.1038/nature16961.
145. ^ "A Google DeepMind Algorithm Uses Deep Learning and More to Master the Game of Go | MIT Technology Review". MIT Technology Review. Retrieved 2016-01-30.
146. ^
147. ^ Marcus, Gary (November 25, 2012). "Is “Deep Learning” a Revolution in Artificial Intelligence?". The New Yorker. Retrieved 2017-06-14.
148. ^ Smith, G. W. (March 27, 2015). "Art and Artificial Intelligence". ArtEnt. Retrieved March 27, 2015.
149. ^ Mellars, Paul (February 1, 2005). "The Impossible Coincidence: A Single-Species Model for the Origins of Modern Human Behavior in Europe" (PDF). Evolutionary Anthropology: Issues, News, and Reviews. Retrieved April 5, 2017.
150. ^ Alexander Mordvintsev; Christopher Olah; Mike Tyka (June 17, 2015). "Inceptionism: Going Deeper into Neural Networks". Google Research Blog. Retrieved June 20, 2015.
151. ^ Alex Hern (June 18, 2015). "Yes, androids do dream of electric sheep". The Guardian. Retrieved June 20, 2015.
152. ^ a b c Goertzel, Ben (2015). "Are there Deep Reasons Underlying the Pathologies of Today's Deep Learning Algorithms?" (PDF).
153. ^ Nguyen, Anh, Jason Yosinski, and Jeff Clune. "Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images." arXiv:1412.1897 (2014).
154. ^ Szegedy, Christian, et al. "Intriguing properties of neural networks." arXiv:1312.6199 (2013).
155. ^ Zhu, S.C.; Mumford, D. "A stochastic grammar of images". Found. Trends Comput. Graph. Vis. 2 (4): 259–362. doi:10.1561/0600000018.
156. ^ Miller, G. A., and N. Chomsky. "Pattern conception." Paper for Conference on pattern detection, University of Michigan. 1957.
157. ^ Eisner, Jason. "Deep Learning of Recursive Structure: Grammar Induction".