DOCSLIB.ORG

RNN and Autoencoder

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
RNN and Autoencoder CS273B lecture 5: RNN and autoencoder James Zou October 10, 2016 Recap Recap: feedforward and convnets Main take-aways: • Composition. Units/layers of a NN are modular and can be composed to form complex architecture. • Weight-sharing. Enforcing that the weight be equal across a set of units can dramatically decrease # of parameters. What are limitations of convnets? What are limitations of convnets? • Fixed input length. • Unclear how to adapt to time-series data. • Convolution corresponds to strong prior—not appropriate for many biological settings. • Could require many labeled training examples (high sample complexity). What are limitations of convnets? • Fixed input length. What are limitations of convnets? • Fixed input length. Recurrent neural network output hidden units input Recurrent neural network output = + · hidden units = ( + ) · input Recurrent neural network output hidden units input Recurrent neural network + − + − + − Recurrent neural network − = ( + + ) − · · − − Recurrent neural network = + − · − − What does RNN remind you of? + − + − + − Vanilla RNN: lacks long term memory + − + − + − LSTM network − − − LSTM network − − − Hochreiter, Schmidhuber 1997 LSTM: inside the hood memory − − Figure adapted from Olah blog. LSTM: inside the hood = + − ∗ = σ( [ , ]+) · − − − Figure adapted from Olah blog. LSTM: inside the hood = + − ∗ ∗ = tanh( [ , ]+) · − = σ( [ , ]+) · − − − Figure adapted from Olah blog. LSTM: inside the hood = σ( [ , ]+) · − = tanh( ) output ∗ − − Figure adapted from Olah blog. LSTM summary • LSTM is a variant of RNN that makes it easier to retain long- range interactions. • Parameters of LSTM: , forget , new memory , weight of new memory (input) , output LSTM application: enhancer/TF prediction Output: 919 binary vector for the presence of TF/chromatin Bi-directional LSTM Similar convolutional architecture as before Input: 200bp sequence Quang and Xie. DanQ. 2016 Deep supervised learning • Feedforward Learning a nonlinear mapping from inputs to outputs. Predicting: • Convnets TF binding, gene expression, disease status from images, • RNN, LSTM risk from SNPs, protein structure … Deep unsupervised learning • Nonlinear dimensional reduction and patterns mining. • In many settings, have more unlabeled examples than labeled. • Learn useful representations from unlabeled data. • Better representation may improve prediction accuracy. Low dimensional structure What is the latent dimensionality of each row of images? Urtasun and Zemel. Autoencoder ˆ decoding () encoding Autoencoder ˆ ˆ = ( + ) · () = ( + ) · , = arg min ˆ , || − || Train with backprop as before. Autoencoder ˆ If encoding and decoding are linear then , = arg min , || − || () What does this remind you of? Autoencoder ˆ If encoding and decoding are linear then , = arg min , || − || () Linear autoencoder is basically just PCA! General f and g corresponds to nonlinear dimensional reduction. What is wrong with this picture? ˆ () What is wrong with this picture? ˆ h(X) can just copy X exactly! Overcomplete. Need to impose sparsity on h. () Denoising autoencoder ˆ () 0 0 independent noise Denoising autoencoder ˆ () 0 0 independent noise Illustration of denoising autoencoder Figure from Hugo Larochelle Filters from denoising autoencoder Basis learned by Basis learned by weight- denoising autoencoder decay autoencoder Deep autoencoder ˆ ˆ Deep autoencoder example original DAE PCA Hinton and Salakhutdinov. Science. 2016 Deep autoencoder example PCA Deep autoencoder Hinton and Salakhutdinov. Science. 2016 Application: deep patient Each patient = vector of 41k clinical descriptors Stack of 3 denoising autoencoder 500 dim representation of each patient Miotto et al. DeepPatient. 2016 Application: deep patient 500 dim representation of each patient Random forest to predict future disease Miotto et al. DeepPatient. 2016 .
Load more

Recommended publications
  • Chombining Recurrent Neural Networks and Adversarial Training
    Combining Recurrent Neural Networks and Adversarial Training for Human Motion Modelling, Synthesis and Control † ‡ Zhiyong Wang∗ Jinxiang Chai Shihong Xia Institute of Computing Texas A&M University Institute of Computing Technology CAS Technology CAS University of Chinese Academy of Sciences ABSTRACT motions from the generator using recurrent neural networks (RNNs) This paper introduces a new generative deep learning network for and refines the generated motion using an adversarial neural network human motion synthesis and control. Our key idea is to combine re- which we call the “refiner network”. Fig 2 gives an overview of current neural networks (RNNs) and adversarial training for human our method: a motion sequence XRNN is generated with the gener- motion modeling. We first describe an efficient method for training ator G and is refined using the refiner network R. To add realism, a RNNs model from prerecorded motion data. We implement recur- we train our refiner network using an adversarial loss, similar to rent neural networks with long short-term memory (LSTM) cells Generative Adversarial Networks (GANs) [9] such that the refined because they are capable of handling nonlinear dynamics and long motion sequences Xre fine are indistinguishable from real motion cap- term temporal dependencies present in human motions. Next, we ture sequences Xreal using a discriminative network D. In addition, train a refiner network using an adversarial loss, similar to Gener- we embed contact information into the generative model to further ative Adversarial Networks (GANs), such that the refined motion improve the quality of the generated motions. sequences are indistinguishable from real motion capture data using We construct the generator G based on recurrent neural network- a discriminative network.
    [Show full text]
  • Training Autoencoders by Alternating Minimization
    Under review as a conference paper at ICLR 2018 TRAINING AUTOENCODERS BY ALTERNATING MINI- MIZATION Anonymous authors Paper under double-blind review ABSTRACT We present DANTE, a novel method for training neural networks, in particular autoencoders, using the alternating minimization principle. DANTE provides a distinct perspective in lieu of traditional gradient-based backpropagation techniques commonly used to train deep networks. It utilizes an adaptation of quasi-convex optimization techniques to cast autoencoder training as a bi-quasi-convex optimiza- tion problem. We show that for autoencoder configurations with both differentiable (e.g. sigmoid) and non-differentiable (e.g. ReLU) activation functions, we can perform the alternations very effectively. DANTE effortlessly extends to networks with multiple hidden layers and varying network configurations. In experiments on standard datasets, autoencoders trained using the proposed method were found to be very promising and competitive to traditional backpropagation techniques, both in terms of quality of solution, as well as training speed. 1 INTRODUCTION For much of the recent march of deep learning, gradient-based backpropagation methods, e.g. Stochastic Gradient Descent (SGD) and its variants, have been the mainstay of practitioners. The use of these methods, especially on vast amounts of data, has led to unprecedented progress in several areas of artificial intelligence. On one hand, the intense focus on these techniques has led to an intimate understanding of hardware requirements and code optimizations needed to execute these routines on large datasets in a scalable manner. Today, myriad off-the-shelf and highly optimized packages exist that can churn reasonably large datasets on GPU architectures with relatively mild human involvement and little bootstrap effort.
    [Show full text]
  • Recurrent Neural Network for Text Classification with Multi-Task
    Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence (IJCAI-16) Recurrent Neural Network for Text Classification with Multi-Task Learning Pengfei Liu Xipeng Qiu⇤ Xuanjing Huang Shanghai Key Laboratory of Intelligent Information Processing, Fudan University School of Computer Science, Fudan University 825 Zhangheng Road, Shanghai, China pfliu14,xpqiu,xjhuang @fudan.edu.cn { } Abstract are based on unsupervised objectives such as word predic- tion for training [Collobert et al., 2011; Turian et al., 2010; Neural network based methods have obtained great Mikolov et al., 2013]. This unsupervised pre-training is effec- progress on a variety of natural language process- tive to improve the final performance, but it does not directly ing tasks. However, in most previous works, the optimize the desired task. models are learned based on single-task super- vised objectives, which often suffer from insuffi- Multi-task learning utilizes the correlation between related cient training data. In this paper, we use the multi- tasks to improve classification by learning tasks in parallel. [ task learning framework to jointly learn across mul- Motivated by the success of multi-task learning Caruana, ] tiple related tasks. Based on recurrent neural net- 1997 , there are several neural network based NLP models [ ] work, we propose three different mechanisms of Collobert and Weston, 2008; Liu et al., 2015b utilize multi- sharing information to model text with task-specific task learning to jointly learn several tasks with the aim of and shared layers. The entire network is trained mutual benefit. The basic multi-task architectures of these jointly on all these tasks. Experiments on four models are to share some lower layers to determine common benchmark text classification tasks show that our features.
    [Show full text]
  • Turbo Autoencoder: Deep Learning Based Channel Codes for Point-To-Point Communication Channels
    Turbo Autoencoder: Deep learning based channel codes for point-to-point communication channels Hyeji Kim Himanshu Asnani Yihan Jiang Samsung AI Center School of Technology ECE Department Cambridge and Computer Science University of Washington Cambridge, United Tata Institute of Seattle, United States Kingdom Fundamental Research [email protected] [email protected] Mumbai, India [email protected] Sewoong Oh Pramod Viswanath Sreeram Kannan Allen School of ECE Department ECE Department Computer Science & University of Illinois at University of Washington Engineering Urbana Champaign Seattle, United States University of Washington Illinois, United States [email protected] Seattle, United States [email protected] [email protected] Abstract Designing codes that combat the noise in a communication medium has remained a significant area of research in information theory as well as wireless communica- tions. Asymptotically optimal channel codes have been developed by mathemati- cians for communicating under canonical models after over 60 years of research. On the other hand, in many non-canonical channel settings, optimal codes do not exist and the codes designed for canonical models are adapted via heuristics to these channels and are thus not guaranteed to be optimal. In this work, we make significant progress on this problem by designing a fully end-to-end jointly trained neural encoder and decoder, namely, Turbo Autoencoder (TurboAE), with the following contributions: (a) under moderate block lengths, TurboAE approaches state-of-the-art performance under canonical channels; (b) moreover, TurboAE outperforms the state-of-the-art codes under non-canonical settings in terms of reliability. TurboAE shows that the development of channel coding design can be automated via deep learning, with near-optimal performance.
    [Show full text]
  • Double Backpropagation for Training Autoencoders Against Adversarial Attack
    1 Double Backpropagation for Training Autoencoders against Adversarial Attack Chengjin Sun, Sizhe Chen, and Xiaolin Huang, Senior Member, IEEE Abstract—Deep learning, as widely known, is vulnerable to adversarial samples. This paper focuses on the adversarial attack on autoencoders. Safety of the autoencoders (AEs) is important because they are widely used as a compression scheme for data storage and transmission, however, the current autoencoders are easily attacked, i.e., one can slightly modify an input but has totally different codes. The vulnerability is rooted the sensitivity of the autoencoders and to enhance the robustness, we propose to adopt double backpropagation (DBP) to secure autoencoder such as VAE and DRAW. We restrict the gradient from the reconstruction image to the original one so that the autoencoder is not sensitive to trivial perturbation produced by the adversarial attack. After smoothing the gradient by DBP, we further smooth the label by Gaussian Mixture Model (GMM), aiming for accurate and robust classification. We demonstrate in MNIST, CelebA, SVHN that our method leads to a robust autoencoder resistant to attack and a robust classifier able for image transition and immune to adversarial attack if combined with GMM. Index Terms—double backpropagation, autoencoder, network robustness, GMM. F 1 INTRODUCTION N the past few years, deep neural networks have been feature [9], [10], [11], [12], [13], or network structure [3], [14], I greatly developed and successfully used in a vast of fields, [15]. such as pattern recognition, intelligent robots, automatic Adversarial attack and its defense are revolving around a control, medicine [1]. Despite the great success, researchers small ∆x and a big resulting difference between f(x + ∆x) have found the vulnerability of deep neural networks to and f(x).
    [Show full text]
  • Artificial Intelligence Applied to Electromechanical Monitoring, A
    ARTIFICIAL INTELLIGENCE APPLIED TO ELECTROMECHANICAL MONITORING, A PERFORMANCE ANALYSIS Erasmus project Authors: Staš Osterc Mentor: Dr. Miguel Delgado Prieto, Dr. Francisco Arellano Espitia 1/5/2020 P a g e II ANNEX VI – DECLARACIÓ D’HONOR P a g e II I declare that, the work in this Master Thesis / Degree Thesis (choose one) is completely my own work, no part of this Master Thesis / Degree Thesis (choose one) is taken from other people’s work without giving them credit, all references have been clearly cited, I’m authorised to make use of the company’s / research group (choose one) related information I’m providing in this document (select when it applies). I understand that an infringement of this declaration leaves me subject to the foreseen disciplinary actions by The Universitat Politècnica de Catalunya - BarcelonaTECH. ___________________ __________________ ___________ Student Name Signature Date Title of the Thesis : _________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ P a g e II Contents Introduction........................................................................................................................................5 Abstract ..........................................................................................................................................5 Aim .................................................................................................................................................6
    [Show full text]
  • Deep Learning Architectures for Sequence Processing
    Speech and Language Processing. Daniel Jurafsky & James H. Martin. Copyright © 2021. All rights reserved. Draft of September 21, 2021. CHAPTER Deep Learning Architectures 9 for Sequence Processing Time will explain. Jane Austen, Persuasion Language is an inherently temporal phenomenon. Spoken language is a sequence of acoustic events over time, and we comprehend and produce both spoken and written language as a continuous input stream. The temporal nature of language is reflected in the metaphors we use; we talk of the flow of conversations, news feeds, and twitter streams, all of which emphasize that language is a sequence that unfolds in time. This temporal nature is reflected in some of the algorithms we use to process lan- guage. For example, the Viterbi algorithm applied to HMM part-of-speech tagging, proceeds through the input a word at a time, carrying forward information gleaned along the way. Yet other machine learning approaches, like those we’ve studied for sentiment analysis or other text classification tasks don’t have this temporal nature – they assume simultaneous access to all aspects of their input. The feedforward networks of Chapter 7 also assumed simultaneous access, al- though they also had a simple model for time. Recall that we applied feedforward networks to language modeling by having them look only at a fixed-size window of words, and then sliding this window over the input, making independent predictions along the way. Fig. 9.1, reproduced from Chapter 7, shows a neural language model with window size 3 predicting what word follows the input for all the. Subsequent words are predicted by sliding the window forward a word at a time.
    [Show full text]
  • Unsupervised Speech Representation Learning Using Wavenet Autoencoders Jan Chorowski, Ron J
    1 Unsupervised speech representation learning using WaveNet autoencoders Jan Chorowski, Ron J. Weiss, Samy Bengio, Aaron¨ van den Oord Abstract—We consider the task of unsupervised extraction speaker gender and identity, from phonetic content, properties of meaningful latent representations of speech by applying which are consistent with internal representations learned autoencoding neural networks to speech waveforms. The goal by speech recognizers [13], [14]. Such representations are is to learn a representation able to capture high level semantic content from the signal, e.g. phoneme identities, while being desired in several tasks, such as low resource automatic speech invariant to confounding low level details in the signal such as recognition (ASR), where only a small amount of labeled the underlying pitch contour or background noise. Since the training data is available. In such scenario, limited amounts learned representation is tuned to contain only phonetic content, of data may be sufficient to learn an acoustic model on the we resort to using a high capacity WaveNet decoder to infer representation discovered without supervision, but insufficient information discarded by the encoder from previous samples. Moreover, the behavior of autoencoder models depends on the to learn the acoustic model and a data representation in a fully kind of constraint that is applied to the latent representation. supervised manner [15], [16]. We compare three variants: a simple dimensionality reduction We focus on representations learned with autoencoders bottleneck, a Gaussian Variational Autoencoder (VAE), and a applied to raw waveforms and spectrogram features and discrete Vector Quantized VAE (VQ-VAE). We analyze the quality investigate the quality of learned representations on LibriSpeech of learned representations in terms of speaker independence, the ability to predict phonetic content, and the ability to accurately re- [17].
    [Show full text]
  • Lecture 11 Recurrent Neural Networks I CMSC 35246: Deep Learning
    Lecture 11 Recurrent Neural Networks I CMSC 35246: Deep Learning Shubhendu Trivedi & Risi Kondor University of Chicago May 01, 2017 Lecture 11 Recurrent Neural Networks I CMSC 35246 Introduction Sequence Learning with Neural Networks Lecture 11 Recurrent Neural Networks I CMSC 35246 Some Sequence Tasks Figure credit: Andrej Karpathy Lecture 11 Recurrent Neural Networks I CMSC 35246 MLPs only accept an input of fixed dimensionality and map it to an output of fixed dimensionality Great e.g.: Inputs - Images, Output - Categories Bad e.g.: Inputs - Text in one language, Output - Text in another language MLPs treat every example independently. How is this problematic? Need to re-learn the rules of language from scratch each time Another example: Classify events after a fixed number of frames in a movie Need to resuse knowledge about the previous events to help in classifying the current. Problems with MLPs for Sequence Tasks The "API" is too limited. Lecture 11 Recurrent Neural Networks I CMSC 35246 Great e.g.: Inputs - Images, Output - Categories Bad e.g.: Inputs - Text in one language, Output - Text in another language MLPs treat every example independently. How is this problematic? Need to re-learn the rules of language from scratch each time Another example: Classify events after a fixed number of frames in a movie Need to resuse knowledge about the previous events to help in classifying the current. Problems with MLPs for Sequence Tasks The "API" is too limited. MLPs only accept an input of fixed dimensionality and map it to an output of fixed dimensionality Lecture 11 Recurrent Neural Networks I CMSC 35246 Bad e.g.: Inputs - Text in one language, Output - Text in another language MLPs treat every example independently.
    [Show full text]
  • Comparative Analysis of Recurrent Neural Network Architectures for Reservoir Inflow Forecasting
    water Article Comparative Analysis of Recurrent Neural Network Architectures for Reservoir Inflow Forecasting Halit Apaydin 1 , Hajar Feizi 2 , Mohammad Taghi Sattari 1,2,* , Muslume Sevba Colak 1 , Shahaboddin Shamshirband 3,4,* and Kwok-Wing Chau 5 1 Department of Agricultural Engineering, Faculty of Agriculture, Ankara University, Ankara 06110, Turkey; [email protected] (H.A.); [email protected] (M.S.C.) 2 Department of Water Engineering, Agriculture Faculty, University of Tabriz, Tabriz 51666, Iran; [email protected] 3 Department for Management of Science and Technology Development, Ton Duc Thang University, Ho Chi Minh City, Vietnam 4 Faculty of Information Technology, Ton Duc Thang University, Ho Chi Minh City, Vietnam 5 Department of Civil and Environmental Engineering, Hong Kong Polytechnic University, Hong Kong, China; [email protected] * Correspondence: [email protected] or [email protected] (M.T.S.); [email protected] (S.S.) Received: 1 April 2020; Accepted: 21 May 2020; Published: 24 May 2020 Abstract: Due to the stochastic nature and complexity of flow, as well as the existence of hydrological uncertainties, predicting streamflow in dam reservoirs, especially in semi-arid and arid areas, is essential for the optimal and timely use of surface water resources. In this research, daily streamflow to the Ermenek hydroelectric dam reservoir located in Turkey is simulated using deep recurrent neural network (RNN) architectures, including bidirectional long short-term memory (Bi-LSTM), gated recurrent unit (GRU), long short-term memory (LSTM), and simple recurrent neural networks (simple RNN). For this purpose, daily observational flow data are used during the period 2012–2018, and all models are coded in Python software programming language.
    [Show full text]
  • A Guide to Recurrent Neural Networks and Backpropagation
    A guide to recurrent neural networks and backpropagation Mikael Bod´en¤ [email protected] School of Information Science, Computer and Electrical Engineering Halmstad University. November 13, 2001 Abstract This paper provides guidance to some of the concepts surrounding recurrent neural networks. Contrary to feedforward networks, recurrent networks can be sensitive, and be adapted to past inputs. Backpropagation learning is described for feedforward networks, adapted to suit our (probabilistic) modeling needs, and extended to cover recurrent net- works. The aim of this brief paper is to set the scene for applying and understanding recurrent neural networks. 1 Introduction It is well known that conventional feedforward neural networks can be used to approximate any spatially finite function given a (potentially very large) set of hidden nodes. That is, for functions which have a fixed input space there is always a way of encoding these functions as neural networks. For a two-layered network, the mapping consists of two steps, y(t) = G(F (x(t))): (1) We can use automatic learning techniques such as backpropagation to find the weights of the network (G and F ) if sufficient samples from the function is available. Recurrent neural networks are fundamentally different from feedforward architectures in the sense that they not only operate on an input space but also on an internal state space – a trace of what already has been processed by the network. This is equivalent to an Iterated Function System (IFS; see (Barnsley, 1993) for a general introduction to IFSs; (Kolen, 1994) for a neural network perspective) or a Dynamical System (DS; see e.g.
    [Show full text]
  • Overview of Lstms and Word2vec and a Bit About Compositional Distributional Semantics If There’S Time
    Overview of LSTMs and word2vec and a bit about compositional distributional semantics if there’s time Ann Copestake Computer Laboratory University of Cambridge November 2016 Outline RNNs and LSTMs Word2vec Compositional distributional semantics Some slides adapted from Aurelie Herbelot. Outline. RNNs and LSTMs Word2vec Compositional distributional semantics Motivation I Standard NNs cannot handle sequence information well. I Can pass them sequences encoded as vectors, but input vectors are fixed length. I Models are needed which are sensitive to sequence input and can output sequences. I RNN: Recurrent neural network. I Long short term memory (LSTM): development of RNN, more effective for most language applications. I More info: http://neuralnetworksanddeeplearning.com/ (mostly about simpler models and CNNs) https://karpathy.github.io/2015/05/21/rnn-effectiveness/ http://colah.github.io/posts/2015-08-Understanding-LSTMs/ Sequences I Video frame categorization: strict time sequence, one output per input. I Real-time speech recognition: strict time sequence. I Neural MT: target not one-to-one with source, order differences. I Many language tasks: best to operate left-to-right and right-to-left (e.g., bi-LSTM). I attention: model ‘concentrates’ on part of input relevant at a particular point. Caption generation: treat image data as ordered, align parts of image with parts of caption. Recurrent Neural Networks http://colah.github.io/posts/ 2015-08-Understanding-LSTMs/ RNN language model: Mikolov et al, 2010 RNN as a language model I Input vector: vector for word at t concatenated to vector which is output from context layer at t − 1. I Performance better than n-grams but won’t capture ‘long-term’ dependencies: She shook her head.
    [Show full text]