Deep learning
TL;DR: Deep learning is making major advances in solving problems that have resisted the best attempts of the artificial intelligence community for many years, and will have many more successes in the near future because it requires very little engineering by hand and can easily take advantage of increases in the amount of available computation and data.
Abstract: Deep learning allows computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction. These methods have dramatically improved the state-of-the-art in speech recognition, visual object recognition, object detection and many other domains such as drug discovery and genomics. Deep learning discovers intricate structure in large data sets by using the backpropagation algorithm to indicate how a machine should change its internal parameters that are used to compute the representation in each layer from the representation in the previous layer. Deep convolutional nets have brought about breakthroughs in processing images, video, speech and audio, whereas recurrent nets have shone light on sequential data such as text and speech.
Citations
More filters
Institute of Cost and Management Accountants of Bangladesh1, Technische Universität München2, Université Paris-Saclay3, Max Planck Society4, University of Pittsburgh5, University of Science and Technology of China6, University of Salerno7, Forschungszentrum Jülich8, Cornell University9, University of Milano-Bicocca10, Universidade Nova de Lisboa11, Kyoto University12, Dresden University of Technology13, Uppsala University14, Nanosystems Initiative Munich15, École Polytechnique Fédérale de Lausanne16, National Institute for Materials Science17, MESA+ Institute for Nanotechnology18, Chalmers University of Technology19, University of Dundee20, Spanish National Research Council21, University of Cambridge22, University of Arkansas23, Polytechnic University of Valencia24, RWTH Aachen University25, Jožef Stefan Institute26
TL;DR: The Towards Oxide-Based Electronics (TO-BE) Action as mentioned in this paper has been recently running in Europe and has involved as participants several hundred scientists from 29 EU countries in a wide four-year project.
251 citations
01 Jan 2018
TL;DR: Interactive COnversational memory Network (ICON), a multi-modal emotion detection framework that extracts multimodal features from conversational videos and hierarchically models the self- and inter-speaker emotional influences into global memories to aid in predicting the emotional orientation of utterance-videos.
Abstract: Emotion recognition in conversations is crucial for building empathetic machines. Present works in this domain do not explicitly consider the inter-personal influences that thrive in the emotional dynamics of dialogues. To this end, we propose Interactive COnversational memory Network (ICON), a multimodal emotion detection framework that extracts multimodal features from conversational videos and hierarchically models the self- and inter-speaker emotional influences into global memories. Such memories generate contextual summaries which aid in predicting the emotional orientation of utterance-videos. Our model outperforms state-of-the-art networks on multiple classification and regression tasks in two benchmark datasets.
251 citations
TL;DR: This editorial addresses both the collection and handling of big data and the analytical tools provided by data science for management scholars, and provides a primer or a “starter kit” for potential data science applications inmanagement research.
Abstract: The recent advent of remote sensing, mobile technologies, novel transaction systems, and highperformance computing offers opportunities to understand trends, behaviors, and actions in a manner that has not been previously possible. Researchers can thus leverage “big data” that are generated from a plurality of sources including mobile transactions, wearable technologies, social media, ambient networks, andbusiness transactions.An earlierAcademy of Management Journal (AMJ) editorial explored the potential implications for data science inmanagement research and highlighted questions for management scholarship as well as the attendant challenges of data sharing and privacy (George, Haas, & Pentland, 2014). This nascent field is evolving rapidly and at a speed that leaves scholars and practitioners alike attempting to make sense of the emergent opportunities that big datahold.With thepromiseof bigdata comequestions about the analytical value and thus relevance of these data for theory development—including concerns over the context-specific relevance, its reliability and its validity. To address this challenge, data science is emerging as an interdisciplinary field that combines statistics, data mining, machine learning, and analytics to understand and explainhowwecan generate analytical insights and prediction models from structured and unstructured big data. Data science emphasizes the systematic study of the organization, properties, and analysis of data and their role in inference, including our confidence in the inference (Dhar, 2013).Whereas both big data and data science terms are often used interchangeably, “big data” refer to large and varied data that can be collected and managed, whereas “data science” develops models that capture, visualize, andanalyze theunderlyingpatterns in thedata. In this editorial, we address both the collection and handling of big data and the analytical tools provided by data science for management scholars. At the current time, practitioners suggest that data science applications tackle the three core elements of big data: volume, velocity, and variety (McAfee & Brynjolfsson, 2012; Zikopoulos & Eaton, 2011). “Volume” represents the sheer size of the dataset due to the aggregation of a large number of variables and an even larger set of observations for each variable. “Velocity” reflects the speed atwhich these data are collected and analyzed, whether in real time or near real time from sensors, sales transactions, social media posts, and sentiment data for breaking news and social trends. “Variety” in big data comes from the plurality of structured and unstructured data sources such as text, videos, networks, and graphics among others. The combinations of volume, velocity, and variety reveal the complex task of generating knowledge from big data, which often runs into millions of observations, and deriving theoretical contributions from such data. In this editorial, we provide a primer or a “starter kit” for potential data science applications inmanagement research. We do so with a caveat that emerging fields outdate and improve uponmethodologies while often supplanting them with new applications. Nevertheless, this primer can guide management scholars who wish to use data science techniques to reach better answers to existing questions or explore completely new research questions.
251 citations
Cites background from "Deep learning"
...Newly introduced approaches such as deep learning (LeCun et al., 2015) can also help overcome the shortcomings of latent Dirichlet allocation (Liu, 2015: 169–171)....
[...]
TL;DR: An overview of AI methods utilised for DR applications is provided, based on a systematic review of over 160 papers, 40 companies and commercial initiatives, and 21 large-scale projects, where AI methods have been used for energy DR.
Abstract: Recent years have seen an increasing interest in Demand Response (DR) as a means to provide flexibility, and hence improve the reliability of energy systems in a cost-effective way. Yet, the high complexity of the tasks associated with DR, combined with their use of large-scale data and the frequent need for near real-time de-cisions, means that Artificial Intelligence (AI) and Machine Learning (ML) — a branch of AI — have recently emerged as key technologies for enabling demand-side response. AI methods can be used to tackle various challenges, ranging from selecting the optimal set of consumers to respond, learning their attributes and pref-erences, dynamic pricing, scheduling and control of devices, learning how to incentivise participants in the DR schemes and how to reward them in a fair and economically efficient way. This work provides an overview of AI methods utilised for DR applications, based on a systematic review of over 160 papers, 40 companies and commercial initiatives, and 21 large-scale projects. The papers are classified with regards to both the AI/ML algorithm(s) used and the application area in energy DR. Next, commercial initiatives are presented (including both start-ups and established companies) and large-scale innovation projects, where AI methods have been used for energy DR. The paper concludes with a discussion of advantages and potential limitations of reviewed AI techniques for different DR tasks, and outlines directions for future research in this fast-growing area.
251 citations
TL;DR: A novel eigenvector based on time–frequency-wavelet joint features is proposed to effectively represent bearing degradation process and a deep autoencoder based joint features compression and computing method is presented to retain effective information without increasing the scale of DNN.
250 citations
References
More filters
TL;DR: A novel, efficient, gradient based method called long short-term memory (LSTM) is introduced, which can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units.
Abstract: Learning to store information over extended time intervals by recurrent backpropagation takes a very long time, mostly because of insufficient, decaying error backflow. We briefly review Hochreiter's (1991) analysis of this problem, then address it by introducing a novel, efficient, gradient based method called long short-term memory (LSTM). Truncating the gradient where this does not do harm, LSTM can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units. Multiplicative gate units learn to open and close access to the constant error flow. LSTM is local in space and time; its computational complexity per time step and weight is O. 1. Our experiments with artificial data involve local, distributed, real-valued, and noisy pattern representations. In comparisons with real-time recurrent learning, back propagation through time, recurrent cascade correlation, Elman nets, and neural sequence chunking, LSTM leads to many more successful runs, and learns much faster. LSTM also solves complex, artificial long-time-lag tasks that have never been solved by previous recurrent network algorithms.
72,897 citations
01 Jan 1998
TL;DR: In this article, a graph transformer network (GTN) is proposed for handwritten character recognition, which can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters.
Abstract: Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.
42,067 citations
TL;DR: Back-propagation repeatedly adjusts the weights of the connections in the network so as to minimize a measure of the difference between the actual output vector of the net and the desired output vector, which helps to represent important features of the task domain.
Abstract: We describe a new learning procedure, back-propagation, for networks of neurone-like units. The procedure repeatedly adjusts the weights of the connections in the network so as to minimize a measure of the difference between the actual output vector of the net and the desired output vector. As a result of the weight adjustments, internal ‘hidden’ units which are not part of the input or output come to represent important features of the task domain, and the regularities in the task are captured by the interactions of these units. The ability to create useful new features distinguishes back-propagation from earlier, simpler methods such as the perceptron-convergence procedure1.
23,814 citations
TL;DR: This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.
Abstract: The theory of reinforcement learning provides a normative account, deeply rooted in psychological and neuroscientific perspectives on animal behaviour, of how agents may optimize their control of an environment. To use reinforcement learning successfully in situations approaching real-world complexity, however, agents are confronted with a difficult task: they must derive efficient representations of the environment from high-dimensional sensory inputs, and use these to generalize past experience to new situations. Remarkably, humans and other animals seem to solve this problem through a harmonious combination of reinforcement learning and hierarchical sensory processing systems, the former evidenced by a wealth of neural data revealing notable parallels between the phasic signals emitted by dopaminergic neurons and temporal difference reinforcement learning algorithms. While reinforcement learning agents have achieved some successes in a variety of domains, their applicability has previously been limited to domains in which useful features can be handcrafted, or to domains with fully observed, low-dimensional state spaces. Here we use recent advances in training deep neural networks to develop a novel artificial agent, termed a deep Q-network, that can learn successful policies directly from high-dimensional sensory inputs using end-to-end reinforcement learning. We tested this agent on the challenging domain of classic Atari 2600 games. We demonstrate that the deep Q-network agent, receiving only the pixels and the game score as inputs, was able to surpass the performance of all previous algorithms and achieve a level comparable to that of a professional human games tester across a set of 49 games, using the same algorithm, network architecture and hyperparameters. This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.
23,074 citations
TL;DR: In this article, an effective way of initializing the weights that allows deep autoencoder networks to learn low-dimensional codes that work much better than principal components analysis as a tool to reduce the dimensionality of data is described.
Abstract: High-dimensional data can be converted to low-dimensional codes by training a multilayer neural network with a small central layer to reconstruct high-dimensional input vectors. Gradient descent can be used for fine-tuning the weights in such "autoencoder" networks, but this works well only if the initial weights are close to a good solution. We describe an effective way of initializing the weights that allows deep autoencoder networks to learn low-dimensional codes that work much better than principal components analysis as a tool to reduce the dimensionality of data.
16,717 citations