On Training of Deep Neural Network. Lornechen

Size: px
Start display at page:

Download "On Training of Deep Neural Network. Lornechen"

Transcription

1 On Training of Deep Neural Network Lornechen

2 Outline Introduction Layer-wise Pre-training & Fine-tuning Activation Function Initialization Method Advanced Layers and Nets 2

3 Neural Network Components Layer: input layer + hidden layer + output layer Neuron: # in each layer and activation function Objective function: LogLoss, CE, RMSE Optimization method: SGD, Nesterov, AdaDelta 3

4 Difficulties of Training DNN Availability of data labeled data is often scarce prone to overfitting Highly non-convex full of bad local optima gradient descent no longer work well Diffusion of gradient the gradients that are propagated backwards rapidly diminish in magnitude as the depth of the network increases. 4

5 Difficulties of Training DNN NN of 2-3 hidden layers are commonly used in the past. However, NN of ten to hundred even to thousand hidden layers are (widely) used nowadays. How? 5

6 Outline Introduction Layer-wise Pre-training & Fine-tuning Activation Function Initialization Method Advanced Layers and Nets 6

7 Supervised Version First, train one layer at a time Fix Fix Train 7

8 Supervised Version First, train one layer at a time Fix Train Fix 8

9 Supervised Version First, train one layer at a time Train Fix Fix 9

10 Supervised Version Finally, fine-tune the whole network BP Forward 10

11 Unsupervised Version 11

12 The Reasoning Pre-training offers a better local optima as a start point compared to randomized initialization. Fine-tuning subsequently adjusts the weights according to the task at hand (guided with objective function). 12

13 However Layer-wise pre-training is rarely used nowadays: requires more training time may lead to a poorer local optimum Just train the network as whole from scratch. 13

14 Outline Introduction Layer-wise Pre-training & Fine-tuning Activation Function Initialization Method Advanced Layers and Nets 14

15 Traditional Activation Function Sigmoid f x = 1 1+e x saturate and kill gradients Tanh f x = ex e x e x +e x similar as sigmoid 15

16 Recap of Back Propagation (BP) dj(w) dw ij (l) = δ i (l+1) f(zj (l) ) δ i (l) = ( j=1 W ji (l) δj (l+1) ) f (z j (l) ) δ i (l) : error of l-th layer W ji (l) : weight between l-th and (l+1)-th layers f(z j (l) ): activation of l-th layer f (z j (l) ): derivative of l-th layer 16

17 Traditional Activation Function f (z j (l) ) < 1 : gradients diminish rapidly when propagated backwards (contractive everywhere) 17

18 Rectified Linear Unit (ReLU) ReLU f x = max(0, x) pros: faster convergence minimum computation cost cons "dead" units 18

19 ReLU Family Leaky ReLU f x = max 0, x + a min(0, x) a is fixed at (very) small value, e.g., 0.01 or 0.3 Parametric ReLU similar as Leaky ReLU but a is learnable Randomized Leaky ReLU similar as Leaky ReLU but a is sampled from a uniform distribution U(l, u) at training, while being fixed at (l+u)/2 at prediction 19

20 ReLU Family 20

21 Exponential Linear Unit (ELU) ELU f x = max 0, x + a min(0, exp x 1) a controls the value to which an ELU saturates for negative net inputs 21

22 What Neuron Type Should I Use? Use the ReLU, be careful with your learning rates and possibly monitor the fraction of "dead" units in a network. If this concerns you, give Leaky ReLU or Maxout a try. Never use sigmoid. Try tanh, but expect it to work worse than ReLU or Maxout. 22

23 Outline Introduction Layer-wise Pre-training & Fine-tuning Activation Function Initialization Method Advanced Layers and Nets 23

24 Traditional Filler Uniform Filler U(-w, w) Gaussian Filler N(0, std), e.g., std = 0.01 difficult to converge for #layers > 8 24

25 Traditional Filler Xavier Filler investigate the variance of responses in each layer assuming linear activation function N(0, std), where std = sqrt(1/n), n = (fan-in + fan-out) / 2 difficult to converge for #layers > 30 (stall) 25

26 MSRA Filler Problem of Xavier Filler assuming linear activation function invalid for ReLU and PReLU New derived filler a = 0, it becomes the ReLU case a = 1, it becomes the linear (the same as Xavier) 26

27 MSRA Filler vs. Xavier Filler Xavier s std will be sqrt(1/2^l) of MSRA Filler if there are L layers. So the signal will be diminishing. Xavier is difficult to converge for #layers >

28 LSUV Filler 28

29 Outline Introduction Layer-wise Pre-training & Fine-tuning Activation Function Initialization Method Advanced Layers and Nets 29

30 Intermediate Supervised Layer overall loss = a * loss0 + b * loss1 + c * loss2 30

31 Deeply Supervised Nets overall loss = a * loss0 + b * loss1 + c * loss2 31

32 Batch Normalization Layer 32

33 Batch Normalization Layer 33

34 Deep Residual Nets Is learning better networks as easy as stacking more layers? a deeper model should produce no higher training error than its shallower counterpart 34

35 Deep Residual Nets Residual Learning Reformation The desired underlying mapping: H(x) Let the stacked nonlinear fit another mapping of F(x) := H(x) x, i.e., residual The original mapping (H) is recast into F(x) + x 35

36 Deep Residual Nets 36

37 Deep Residual Nets 37

38 Reference Xavier Filler: Understanding the Difficulty of Training Deep Feedforward Neural Networks PReLU & MSRA Filler: Delving Deep into Rectifiers Surpassing Human- Level Performance on ImageNet Classification RReLU: Empirical Evaluation of Rectified Activations in Convolution Network ELU: Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs) LSUV Filler: All You Need Is A Good Init GoogleNet: Going Deeper with Convolutions DSN: Deeply-Supervised Nets BatchNorm: Batch Normalization: Accelerating Deep Network Training b y Reducing Internal Covariate Shift ResNet: Deep Residual Learning for Image Recognition 38

39 Thanks 39

CSE Introduction to High-Perfomance Deep Learning ImageNet & VGG. Jihyung Kil

CSE Introduction to High-Perfomance Deep Learning ImageNet & VGG. Jihyung Kil CSE 5194.01 - Introduction to High-Perfomance Deep Learning ImageNet & VGG Jihyung Kil ImageNet Classification with Deep Convolutional Neural Networks Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton,

More information

COMP9444 Neural Networks and Deep Learning 5. Convolutional Networks

COMP9444 Neural Networks and Deep Learning 5. Convolutional Networks COMP9444 Neural Networks and Deep Learning 5. Convolutional Networks Textbook, Sections 6.2.2, 6.3, 7.9, 7.11-7.13, 9.1-9.5 COMP9444 17s2 Convolutional Networks 1 Outline Geometry of Hidden Unit Activations

More information

Retinopathy Net. Alberto Benavides Robert Dadashi Neel Vadoothker

Retinopathy Net. Alberto Benavides Robert Dadashi Neel Vadoothker Retinopathy Net Alberto Benavides Robert Dadashi Neel Vadoothker Motivation We were interested in applying deep learning techniques to the field of medical imaging Field holds a lot of promise and can

More information

Summary and discussion of: Why Does Unsupervised Pre-training Help Deep Learning?

Summary and discussion of: Why Does Unsupervised Pre-training Help Deep Learning? Summary and discussion of: Why Does Unsupervised Pre-training Help Deep Learning? Statistics Journal Club, 36-825 Avinava Dubey and Mrinmaya Sachan and Jerzy Wieczorek December 3, 2014 1 Summary 1.1 Deep

More information

CS 453X: Class 18. Jacob Whitehill

CS 453X: Class 18. Jacob Whitehill CS 453X: Class 18 Jacob Whitehill More on k-means Exercise: Empty clusters (1) Assume that a set of distinct data points { x (i) } are initially assigned so that none of the k clusters is empty. How can

More information

Reduction of Overfitting in Diabetes Prediction Using Deep Learning Neural Network

Reduction of Overfitting in Diabetes Prediction Using Deep Learning Neural Network Reduction of Overfitting in Diabetes Prediction Using Deep Learning Neural Network Akm Ashiquzzaman *, Abdul Kawsar Tushar *, Md. Rashedul Islam *, 1, and Jong-Myon Kim **, 2 * Department of CSE, University

More information

Efficient Deep Model Selection

Efficient Deep Model Selection Efficient Deep Model Selection Jose Alvarez Researcher Data61, CSIRO, Australia GTC, May 9 th 2017 www.josemalvarez.net conv1 conv2 conv3 conv4 conv5 conv6 conv7 conv8 softmax prediction???????? Num Classes

More information

Cost-aware Pre-training for Multiclass Cost-sensitive Deep Learning

Cost-aware Pre-training for Multiclass Cost-sensitive Deep Learning Cost-aware Pre-training for Multiclass Cost-sensitive Deep Learning Yu-An Chung 1 Hsuan-Tien Lin 1 Shao-Wen Yang 2 1 Dept. of Computer Science and Information Engineering National Taiwan University, Taiwan

More information

Multi-attention Guided Activation Propagation in CNNs

Multi-attention Guided Activation Propagation in CNNs Multi-attention Guided Activation Propagation in CNNs Xiangteng He and Yuxin Peng (B) Institute of Computer Science and Technology, Peking University, Beijing, China pengyuxin@pku.edu.cn Abstract. CNNs

More information

Factoid Question Answering

Factoid Question Answering Factoid Question Answering CS 898 Project June 12, 2017 Salman Mohammed David R. Cheriton School of Computer Science University of Waterloo Motivation Source: https://www.apple.com/newsroom/2017/01/hey-siri-whos-going-to-win-the-super-bowl/

More information

Learning in neural networks

Learning in neural networks http://ccnl.psy.unipd.it Learning in neural networks Marco Zorzi University of Padova M. Zorzi - European Diploma in Cognitive and Brain Sciences, Cognitive modeling", HWK 19-24/3/2006 1 Connectionist

More information

Intelligent Systems. Discriminative Learning. Parts marked by * are optional. WS2013/2014 Carsten Rother, Dmitrij Schlesinger

Intelligent Systems. Discriminative Learning. Parts marked by * are optional. WS2013/2014 Carsten Rother, Dmitrij Schlesinger Intelligent Systems Discriminative Learning Parts marked by * are optional 30/12/2013 WS2013/2014 Carsten Rother, Dmitrij Schlesinger Discriminative models There exists a joint probability distribution

More information

The Impact of Visual Saliency Prediction in Image Classification

The Impact of Visual Saliency Prediction in Image Classification Dublin City University Insight Centre for Data Analytics Universitat Politecnica de Catalunya Escola Tècnica Superior d Enginyeria de Telecomunicacions de Barcelona Eric Arazo Sánchez The Impact of Visual

More information

Neurons and neural networks II. Hopfield network

Neurons and neural networks II. Hopfield network Neurons and neural networks II. Hopfield network 1 Perceptron recap key ingredient: adaptivity of the system unsupervised vs supervised learning architecture for discrimination: single neuron perceptron

More information

Introduction to Machine Learning. Katherine Heller Deep Learning Summer School 2018

Introduction to Machine Learning. Katherine Heller Deep Learning Summer School 2018 Introduction to Machine Learning Katherine Heller Deep Learning Summer School 2018 Outline Kinds of machine learning Linear regression Regularization Bayesian methods Logistic Regression Why we do this

More information

A CONVOLUTION NEURAL NETWORK ALGORITHM FOR BRAIN TUMOR IMAGE SEGMENTATION

A CONVOLUTION NEURAL NETWORK ALGORITHM FOR BRAIN TUMOR IMAGE SEGMENTATION A CONVOLUTION NEURAL NETWORK ALGORITHM FOR BRAIN TUMOR IMAGE SEGMENTATION 1 Priya K, 2 Dr.O. Saraniya 1 PG Scholar, 2 Assistant Professor Department Of ECE Government College of Technology, Coimbatore,

More information

Convolutional Neural Networks for Text Classification

Convolutional Neural Networks for Text Classification Convolutional Neural Networks for Text Classification Sebastian Sierra MindLab Research Group July 1, 2016 ebastian Sierra (MindLab Research Group) NLP Summer Class July 1, 2016 1 / 32 Outline 1 What is

More information

Understanding and Improving Deep Learning Algorithms

Understanding and Improving Deep Learning Algorithms Understanding and Improving Deep Learning Algorithms Yoshua Bengio, U. Montreal CMU, Pittsburgh April 22nd, 2010 ML Google Distinguished Lecture Thanks to: Aaron Courville, Xavier Glorot, Dumitru Erhan,

More information

Flexible, High Performance Convolutional Neural Networks for Image Classification

Flexible, High Performance Convolutional Neural Networks for Image Classification Flexible, High Performance Convolutional Neural Networks for Image Classification Dan C. Cireşan, Ueli Meier, Jonathan Masci, Luca M. Gambardella, Jürgen Schmidhuber IDSIA, USI and SUPSI Manno-Lugano,

More information

PMR5406 Redes Neurais e Lógica Fuzzy. Aula 5 Alguns Exemplos

PMR5406 Redes Neurais e Lógica Fuzzy. Aula 5 Alguns Exemplos PMR5406 Redes Neurais e Lógica Fuzzy Aula 5 Alguns Exemplos APPLICATIONS Two examples of real life applications of neural networks for pattern classification: RBF networks for face recognition FF networks

More information

DIABETIC RISK PREDICTION FOR WOMEN USING BOOTSTRAP AGGREGATION ON BACK-PROPAGATION NEURAL NETWORKS

DIABETIC RISK PREDICTION FOR WOMEN USING BOOTSTRAP AGGREGATION ON BACK-PROPAGATION NEURAL NETWORKS International Journal of Computer Engineering & Technology (IJCET) Volume 9, Issue 4, July-Aug 2018, pp. 196-201, Article IJCET_09_04_021 Available online at http://www.iaeme.com/ijcet/issues.asp?jtype=ijcet&vtype=9&itype=4

More information

arxiv: v2 [cs.cv] 7 Jun 2018

arxiv: v2 [cs.cv] 7 Jun 2018 Deep supervision with additional labels for retinal vessel segmentation task Yishuo Zhang and Albert C.S. Chung Lo Kwee-Seong Medical Image Analysis Laboratory, Department of Computer Science and Engineering,

More information

Motivation: Attention: Focusing on specific parts of the input. Inspired by neuroscience.

Motivation: Attention: Focusing on specific parts of the input. Inspired by neuroscience. Outline: Motivation. What s the attention mechanism? Soft attention vs. Hard attention. Attention in Machine translation. Attention in Image captioning. State-of-the-art. 1 Motivation: Attention: Focusing

More information

Elad Hoffer*, Itay Hubara*, Daniel Soudry

Elad Hoffer*, Itay Hubara*, Daniel Soudry Train longer, generalize better: closing the generalization gap in large batch training of neural networks Elad Hoffer*, Itay Hubara*, Daniel Soudry *Equal contribution Better models - parallelization

More information

Convolutional Neural Networks for Estimating Left Ventricular Volume

Convolutional Neural Networks for Estimating Left Ventricular Volume Convolutional Neural Networks for Estimating Left Ventricular Volume Ryan Silva Stanford University rdsilva@stanford.edu Maksim Korolev Stanford University mkorolev@stanford.edu Abstract End-systolic and

More information

Large-Batch Training and Generalization Gap. Nitish Shirish Keskar Salesforce Research Slides: keskarnitish.github.io

Large-Batch Training and Generalization Gap. Nitish Shirish Keskar Salesforce Research Slides: keskarnitish.github.io Large-Batch Training and Generalization Gap Nitish Shirish Keskar Salesforce Research Slides: keskarnitish.github.io 1 Deep Learning & SGD 2 Deep Learning & SGD 2 Deep Learning & SGD Large batch sizes

More information

Convolutional Neural Networks (CNN)

Convolutional Neural Networks (CNN) Convolutional Neural Networks (CNN) Algorithm and Some Applications in Computer Vision Luo Hengliang Institute of Automation June 10, 2014 Luo Hengliang (Institute of Automation) Convolutional Neural Networks

More information

A Comparison of Deep Neural Network Training Methods for Large Vocabulary Speech Recognition

A Comparison of Deep Neural Network Training Methods for Large Vocabulary Speech Recognition A Comparison of Deep Neural Network Training Methods for Large Vocabulary Speech Recognition LászlóTóth and Tamás Grósz MTA-SZTE Research Group on Artificial Intelligence Hungarian Academy of Sciences

More information

Data mining for Obstructive Sleep Apnea Detection. 18 October 2017 Konstantinos Nikolaidis

Data mining for Obstructive Sleep Apnea Detection. 18 October 2017 Konstantinos Nikolaidis Data mining for Obstructive Sleep Apnea Detection 18 October 2017 Konstantinos Nikolaidis Introduction: What is Obstructive Sleep Apnea? Obstructive Sleep Apnea (OSA) is a relatively common sleep disorder

More information

Training deep Autoencoders for collaborative filtering Oleksii Kuchaiev & Boris Ginsburg

Training deep Autoencoders for collaborative filtering Oleksii Kuchaiev & Boris Ginsburg Training deep Autoencoders for collaborative filtering Oleksii Kuchaiev & Boris Ginsburg Motivation Personalized recommendations 2 Key points (spoiler alert) 1. Deep autoencoder for collaborative filtering

More information

Understanding Convolutional Neural

Understanding Convolutional Neural Understanding Convolutional Neural Networks Understanding Convolutional Neural Networks David Stutz July 24th, 2014 David Stutz July 24th, 2014 0/53 1/53 Table of Contents - Table of Contents 1 Motivation

More information

TIME SERIES MODELING USING ARTIFICIAL NEURAL NETWORKS 1 P.Ram Kumar, 2 M.V.Ramana Murthy, 3 D.Eashwar, 4 M.Venkatdas

TIME SERIES MODELING USING ARTIFICIAL NEURAL NETWORKS 1 P.Ram Kumar, 2 M.V.Ramana Murthy, 3 D.Eashwar, 4 M.Venkatdas TIME SERIES MODELING USING ARTIFICIAL NEURAL NETWORKS 1 P.Ram Kumar, 2 M.V.Ramana Murthy, 3 D.Eashwar, 4 M.Venkatdas 1 Department of Computer Science & Engineering,UCE,OU,Hyderabad 2 Department of Mathematics,UCS,OU,Hyderabad

More information

UNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Final, Fall 2014

UNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Final, Fall 2014 UNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Final, Fall 2014 Exam policy: This exam allows two one-page, two-sided cheat sheets (i.e. 4 sides); No other materials. Time: 2 hours. Be sure to write

More information

Why did the network make this prediction?

Why did the network make this prediction? Why did the network make this prediction? Ankur Taly (Google Inc.) Joint work with Mukund Sundararajan and Qiqi Yan Some Deep Learning Successes Source: https://www.cbsinsights.com Deep Neural Networks

More information

Review: Logistic regression, Gaussian naïve Bayes, linear regression, and their connections

Review: Logistic regression, Gaussian naïve Bayes, linear regression, and their connections Review: Logistic regression, Gaussian naïve Bayes, linear regression, and their connections New: Bias-variance decomposition, biasvariance tradeoff, overfitting, regularization, and feature selection Yi

More information

Question 1 Multiple Choice (8 marks)

Question 1 Multiple Choice (8 marks) Philadelphia University Student Name: Faculty of Engineering Student Number: Dept. of Computer Engineering First Exam, First Semester: 2015/2016 Course Title: Neural Networks and Fuzzy Logic Date: 19/11/2015

More information

Synthesis of Gadolinium-enhanced MRI for Multiple Sclerosis patients using Generative Adversarial Network

Synthesis of Gadolinium-enhanced MRI for Multiple Sclerosis patients using Generative Adversarial Network Medical Application of GAN Synthesis of Gadolinium-enhanced MRI for Multiple Sclerosis patients using Generative Adversarial Network Sumana Basu School of Computer Science McGill University 260727568 sumana.basu@mail.mcgill.ca

More information

Deep Learning Analytics for Predicting Prognosis of Acute Myeloid Leukemia with Cytogenetics, Age, and Mutations

Deep Learning Analytics for Predicting Prognosis of Acute Myeloid Leukemia with Cytogenetics, Age, and Mutations Deep Learning Analytics for Predicting Prognosis of Acute Myeloid Leukemia with Cytogenetics, Age, and Mutations Andy Nguyen, M.D., M.S. Medical Director, Hematopathology, Hematology and Coagulation Laboratory,

More information

Computational Intelligence Lecture 21: Integrating Fuzzy Systems and Neural Networks

Computational Intelligence Lecture 21: Integrating Fuzzy Systems and Neural Networks Computational Intelligence Lecture 21: Integrating Fuzzy Systems and Neural Networks Farzaneh Abdollahi Department of Electrical Engineering Amirkabir University of Technology Fall 2013 Farzaneh Abdollahi

More information

A HMM-based Pre-training Approach for Sequential Data

A HMM-based Pre-training Approach for Sequential Data A HMM-based Pre-training Approach for Sequential Data Luca Pasa 1, Alberto Testolin 2, Alessandro Sperduti 1 1- Department of Mathematics 2- Department of Developmental Psychology and Socialisation University

More information

Conditional Activation for Diverse Neurons in Heterogeneous Networks

Conditional Activation for Diverse Neurons in Heterogeneous Networks Conditional Activation for Diverse Neurons in Heterogeneous Networks Abstract Albert Lee 1, Bonnie Lam 1, Wenyuan Li 1, Hochul Lee 1, Wei-Hao Chen 2, Meng-Fan Chang 2, and Kang. -L. Wang 1 1 University

More information

CSE 258 Lecture 2. Web Mining and Recommender Systems. Supervised learning Regression

CSE 258 Lecture 2. Web Mining and Recommender Systems. Supervised learning Regression CSE 258 Lecture 2 Web Mining and Recommender Systems Supervised learning Regression Supervised versus unsupervised learning Learning approaches attempt to model data in order to solve a problem Unsupervised

More information

EECS 433 Statistical Pattern Recognition

EECS 433 Statistical Pattern Recognition EECS 433 Statistical Pattern Recognition Ying Wu Electrical Engineering and Computer Science Northwestern University Evanston, IL 60208 http://www.eecs.northwestern.edu/~yingwu 1 / 19 Outline What is Pattern

More information

Learning Convolutional Neural Networks for Graphs

Learning Convolutional Neural Networks for Graphs GA-65449 Learning Convolutional Neural Networks for Graphs Mathias Niepert Mohamed Ahmed Konstantin Kutzkov NEC Laboratories Europe Representation Learning for Graphs Telecom Safety Transportation Industry

More information

STDP-based spiking deep convolutional neural networks for object recognition

STDP-based spiking deep convolutional neural networks for object recognition arxiv:1611.01421v3 [cs.cv] 25 Dec 2017 This manuscript is published in Neural Networks. Please cite it as: Kheradpisheh, S.R., Ganjtabesh, M., Thorpe, S.J., Masquelier, T., STDP-based spiking deep convolutional

More information

Tumor Cellularity Assessment. Rene Bidart

Tumor Cellularity Assessment. Rene Bidart Tumor Cellularity Assessment Rene Bidart 1 Goals Find location and class of all cell nuclei: Lymphocyte Cells Normal Epithelial Cells Malignant Epithelial Cells Why: Test effectiveness of pre-surgery treatment

More information

Classıfıcatıon of Dıabetes Dısease Usıng Backpropagatıon and Radıal Basıs Functıon Network

Classıfıcatıon of Dıabetes Dısease Usıng Backpropagatıon and Radıal Basıs Functıon Network UTM Computing Proceedings Innovations in Computing Technology and Applications Volume 2 Year: 2017 ISBN: 978-967-0194-95-0 1 Classıfıcatıon of Dıabetes Dısease Usıng Backpropagatıon and Radıal Basıs Functıon

More information

arxiv: v2 [cs.lg] 1 Jun 2018

arxiv: v2 [cs.lg] 1 Jun 2018 Shagun Sodhani 1 * Vardaan Pahuja 1 * arxiv:1805.11016v2 [cs.lg] 1 Jun 2018 Abstract Self-play (Sukhbaatar et al., 2017) is an unsupervised training procedure which enables the reinforcement learning agents

More information

Speech Enhancement Using Deep Neural Network

Speech Enhancement Using Deep Neural Network Speech Enhancement Using Deep Neural Network Pallavi D. Bhamre 1, Hemangi H. Kulkarni 2 1 Post-graduate Student, Department of Electronics and Telecommunication, R. H. Sapat College of Engineering, Management

More information

Speech Enhancement Based on Deep Neural Networks

Speech Enhancement Based on Deep Neural Networks Speech Enhancement Based on Deep Neural Networks Chin-Hui Lee School of ECE, Georgia Tech chl@ece.gatech.edu Joint work with Yong Xu and Jun Du at USTC 1 Outline and Talk Agenda In Signal Processing Letter,

More information

Supplementary method. a a a. a a a. h h,w w, then the size of C will be ( h h + 1) ( w w + 1). The matrix K is

Supplementary method. a a a. a a a. h h,w w, then the size of C will be ( h h + 1) ( w w + 1). The matrix K is Deep earning Radiomics of Shear Wave lastography Significantly Improved Diagnostic Performance for Assessing iver Fibrosis in Chronic epatitis B: A Prospective Multicenter Study Supplementary method Mathematical

More information

Object Detectors Emerge in Deep Scene CNNs

Object Detectors Emerge in Deep Scene CNNs Object Detectors Emerge in Deep Scene CNNs Bolei Zhou, Aditya Khosla, Agata Lapedriza, Aude Oliva, Antonio Torralba Presented By: Collin McCarthy Goal: Understand how objects are represented in CNNs Are

More information

Figure 1: MRI Scanning [2]

Figure 1: MRI Scanning [2] A Deep Belief Network Based Brain Tumor Detection in MRI Images Thahseen P 1, Anish Kumar B 2 1 MEA Engineering College, State Highway 39, Nellikunnu-Vengoor, Perinthalmanna, Malappuram, Kerala 2 Assistant

More information

Acoustic Signal Processing Based on Deep Neural Networks

Acoustic Signal Processing Based on Deep Neural Networks Acoustic Signal Processing Based on Deep Neural Networks Chin-Hui Lee School of ECE, Georgia Tech chl@ece.gatech.edu Joint work with Yong Xu, Yanhui Tu, Qing Wang, Tian Gao, Jun Du, LiRong Dai Outline

More information

Deep Learning for Predicting In Hospital Mortality

Deep Learning for Predicting In Hospital Mortality Deep Learning for Predicting In Hospital Mortality CHRISTINE LEE, MS, UC IRVINE BIOMEDICAL ENGINEERING IRA HOFER, MD, UCLA ANESTHESIOLOGY MAXIME CANNESSON, MD PHD, UCLA ANESTHESIOLOGY PIERRE BALDI, PHD,

More information

Using stigmergy to incorporate the time into artificial neural networks

Using stigmergy to incorporate the time into artificial neural networks Using stigmergy to incorporate the time into artificial neural networks Federico A. Galatolo, Mario G.C.A. Cimino, and Gigliola Vaglini Department of Information Engineering, University of Pisa, 56122

More information

Auto-Encoder Pre-Training of Segmented-Memory Recurrent Neural Networks

Auto-Encoder Pre-Training of Segmented-Memory Recurrent Neural Networks Auto-Encoder Pre-Training of Segmented-Memory Recurrent Neural Networks Stefan Glüge, Ronald Böck and Andreas Wendemuth Faculty of Electrical Engineering and Information Technology Cognitive Systems Group,

More information

An Ensemble CNN2ELM for Age Estimation. ,KenliLi,Senior Member, IEEE, and Keqin Li, Fellow, IEEE

An Ensemble CNN2ELM for Age Estimation. ,KenliLi,Senior Member, IEEE, and Keqin Li, Fellow, IEEE 758 IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 13, NO. 3, MARCH 018 An Ensemble CNNELM for Age Estimation Mingxing Duan,KenliLi,Senior Member, IEEE, and Keqin Li, Fellow, IEEE Abstract

More information

arxiv: v1 [cs.ne] 5 Dec 2018

arxiv: v1 [cs.ne] 5 Dec 2018 Neuromodulated Learning in Deep Neural Networks Dennis G Wilson University of Toulouse, IRIT - CNRS - UMR5505, Toulouse, France dennis.wilson@irit.fr arxiv:1812.03365v1 [cs.ne] 5 Dec 2018 Sylvain Cussat-Blanc

More information

Deep Learning of Brain Lesion Patterns for Predicting Future Disease Activity in Patients with Early Symptoms of Multiple Sclerosis

Deep Learning of Brain Lesion Patterns for Predicting Future Disease Activity in Patients with Early Symptoms of Multiple Sclerosis Deep Learning of Brain Lesion Patterns for Predicting Future Disease Activity in Patients with Early Symptoms of Multiple Sclerosis Youngjin Yoo 1,2,5(B),LisaW.Tang 2,3,5, Tom Brosch 1,2,5,DavidK.B.Li

More information

Learning and Adaptive Behavior, Part II

Learning and Adaptive Behavior, Part II Learning and Adaptive Behavior, Part II April 12, 2007 The man who sets out to carry a cat by its tail learns something that will always be useful and which will never grow dim or doubtful. -- Mark Twain

More information

Exploratory Study on Direct Prediction of Diabetes using Deep Residual Networks

Exploratory Study on Direct Prediction of Diabetes using Deep Residual Networks Exploratory Study on Direct Prediction of Diabetes using Deep Residual Networks Samaneh Abbasi-Sureshjani, Behdad Dashtbozorg, Bart M. ter Haar Romeny, and François Fleuret Abstract Diabetes is threatening

More information

Chapter 1. Introduction

Chapter 1. Introduction Chapter 1 Introduction Artificial neural networks are mathematical inventions inspired by observations made in the study of biological systems, though loosely based on the actual biology. An artificial

More information

Keywords Artificial Neural Networks (ANN), Echocardiogram, BPNN, RBFNN, Classification, survival Analysis.

Keywords Artificial Neural Networks (ANN), Echocardiogram, BPNN, RBFNN, Classification, survival Analysis. Design of Classifier Using Artificial Neural Network for Patients Survival Analysis J. D. Dhande 1, Dr. S.M. Gulhane 2 Assistant Professor, BDCE, Sevagram 1, Professor, J.D.I.E.T, Yavatmal 2 Abstract The

More information

Noise Cancellation using Adaptive Filters Algorithms

Noise Cancellation using Adaptive Filters Algorithms Noise Cancellation using Adaptive Filters Algorithms Suman, Poonam Beniwal Department of ECE, OITM, Hisar, bhariasuman13@gmail.com Abstract Active Noise Control (ANC) involves an electro acoustic or electromechanical

More information

Supplementary Online Content

Supplementary Online Content Supplementary Online Content Ting DS, Cheung CY-L, Lim G, et al. Development and validation of a deep learning system for diabetic retinopathy and related eye diseases using retinal images from multiethnic

More information

arxiv: v2 [cs.cv] 22 Mar 2018

arxiv: v2 [cs.cv] 22 Mar 2018 Deep saliency: What is learnt by a deep network about saliency? Sen He 1 Nicolas Pugeault 1 arxiv:1801.04261v2 [cs.cv] 22 Mar 2018 Abstract Deep convolutional neural networks have achieved impressive performance

More information

Aristomenis Kotsakis,Matthias Nübling, Nikolaos P. Bakas, George Pelekanakis, John Thanopoulos

Aristomenis Kotsakis,Matthias Nübling, Nikolaos P. Bakas, George Pelekanakis, John Thanopoulos 2nd International Conference on Sustainable Employability Building Bridges between Science and Practice - http://www.employability21.com/ 12-13 September 2018 Provinciehuis Vlaams Brabant, Leuven, Belgium

More information

Network Dissection: Quantifying Interpretability of Deep Visual Representation

Network Dissection: Quantifying Interpretability of Deep Visual Representation Name: Pingchuan Ma Student number: 3526400 Date: August 19, 2018 Seminar: Explainable Machine Learning Lecturer: PD Dr. Ullrich Köthe SS 2018 Quantifying Interpretability of Deep Visual Representation

More information

Evidence that the ventral stream codes the errors used in hierarchical inference and learning Running title: Error coding in the ventral stream

Evidence that the ventral stream codes the errors used in hierarchical inference and learning Running title: Error coding in the ventral stream 1 2 3 4 5 6 7 8 9 1 11 12 13 14 15 16 17 18 19 2 21 22 23 24 25 26 27 28 29 3 31 32 33 34 35 36 37 38 39 4 41 42 43 44 45 46 47 48 49 Evidence that the ventral stream codes the errors used in hierarchical

More information

Artificial Neural Networks and Near Infrared Spectroscopy - A case study on protein content in whole wheat grain

Artificial Neural Networks and Near Infrared Spectroscopy - A case study on protein content in whole wheat grain A White Paper from FOSS Artificial Neural Networks and Near Infrared Spectroscopy - A case study on protein content in whole wheat grain By Lars Nørgaard*, Martin Lagerholm and Mark Westerhaus, FOSS *corresponding

More information

Artificial Neural Networks (Ref: Negnevitsky, M. Artificial Intelligence, Chapter 6)

Artificial Neural Networks (Ref: Negnevitsky, M. Artificial Intelligence, Chapter 6) Artificial Neural Networks (Ref: Negnevitsky, M. Artificial Intelligence, Chapter 6) BPNN in Practice Week 3 Lecture Notes page 1 of 1 The Hopfield Network In this network, it was designed on analogy of

More information

Automated diagnosis of pneumothorax using an ensemble of convolutional neural networks with multi-sized chest radiography images

Automated diagnosis of pneumothorax using an ensemble of convolutional neural networks with multi-sized chest radiography images Automated diagnosis of pneumothorax using an ensemble of convolutional neural networks with multi-sized chest radiography images Tae Joon Jun, Dohyeun Kim, and Daeyoung Kim School of Computing, KAIST,

More information

Deep Neural Networks: A New Framework for Modeling Biological Vision and Brain Information Processing

Deep Neural Networks: A New Framework for Modeling Biological Vision and Brain Information Processing ANNUAL REVIEWS Further Click here to view this article's online features: Download figures as PPT slides Navigate linked references Download citations Explore related articles Search keywords Annu. Rev.

More information

Task 1: Machine Learning with Spike-Timing-Dependent Plasticity (STDP)

Task 1: Machine Learning with Spike-Timing-Dependent Plasticity (STDP) DARPA Report Task1 for Year 1 (Q1-Q4) Task 1: Machine Learning with Spike-Timing-Dependent Plasticity (STDP) 1. Shortcomings of the deep learning approach to artificial intelligence It has been established

More information

Computational Cognitive Neuroscience

Computational Cognitive Neuroscience Computational Cognitive Neuroscience Computational Cognitive Neuroscience Computational Cognitive Neuroscience *Computer vision, *Pattern recognition, *Classification, *Picking the relevant information

More information

Image Captioning using Reinforcement Learning. Presentation by: Samarth Gupta

Image Captioning using Reinforcement Learning. Presentation by: Samarth Gupta Image Captioning using Reinforcement Learning Presentation by: Samarth Gupta 1 Introduction Summary Supervised Models Image captioning as RL problem Actor Critic Architecture Policy Gradient architecture

More information

Lung Nodule Segmentation Using 3D Convolutional Neural Networks

Lung Nodule Segmentation Using 3D Convolutional Neural Networks Lung Nodule Segmentation Using 3D Convolutional Neural Networks Research paper Business Analytics Bernard Bronmans Master Business Analytics VU University, Amsterdam Evert Haasdijk Supervisor VU University,

More information

Final Report: Automated Semantic Segmentation of Volumetric Cardiovascular Features and Disease Assessment

Final Report: Automated Semantic Segmentation of Volumetric Cardiovascular Features and Disease Assessment Final Report: Automated Semantic Segmentation of Volumetric Cardiovascular Features and Disease Assessment Tony Lindsey 1,3, Xiao Lu 1 and Mojtaba Tefagh 2 1 Department of Biomedical Informatics, Stanford

More information

Katsunari Shibata and Tomohiko Kawano

Katsunari Shibata and Tomohiko Kawano Learning of Action Generation from Raw Camera Images in a Real-World-Like Environment by Simple Coupling of Reinforcement Learning and a Neural Network Katsunari Shibata and Tomohiko Kawano Oita University,

More information

Magnetic Resonance Contrast Prediction Using Deep Learning

Magnetic Resonance Contrast Prediction Using Deep Learning Magnetic Resonance Contrast Prediction Using Deep Learning Cagan Alkan Department of Electrical Engineering Stanford University calkan@stanford.edu Andrew Weitz Department of Bioengineering Stanford University

More information

Interpretable & Transparent Deep Learning

Interpretable & Transparent Deep Learning Fraunhofer Image Processing Heinrich Hertz Institute Interpretable & Transparent Deep Learning Fraunhofer HHI, Machine Learning Group Wojciech Samek Northern Lights Deep Learning Workshop (NLDL 19) Tromsø,

More information

A Novel Method using Convolutional Neural Network for Segmenting Brain Tumor in MRI Images

A Novel Method using Convolutional Neural Network for Segmenting Brain Tumor in MRI Images A Novel Method using Convolutional Neural Network for Segmenting Brain Tumor in MRI Images K. Meena*, S. Pavitra**, N. Nishanthi*** & M. Nivetha**** *UG Student, Department of Electronics and Communication

More information

J2.6 Imputation of missing data with nonlinear relationships

J2.6 Imputation of missing data with nonlinear relationships Sixth Conference on Artificial Intelligence Applications to Environmental Science 88th AMS Annual Meeting, New Orleans, LA 20-24 January 2008 J2.6 Imputation of missing with nonlinear relationships Michael

More information

c Copyright 2017 Cody Burkard

c Copyright 2017 Cody Burkard c Copyright 2017 Cody Burkard Can Intelligent Hyperparameter Selection Improve Resistance to Adversarial Examples? Cody Burkard A thesis submitted in partial fulfillment of the requirements for the degree

More information

Generative Adversarial Networks.

Generative Adversarial Networks. Generative Adversarial Networks www.cs.wisc.edu/~page/cs760/ Goals for the lecture you should understand the following concepts Nash equilibrium Minimax game Generative adversarial network Prisoners Dilemma

More information

Lesson 6 Learning II Anders Lyhne Christensen, D6.05, INTRODUCTION TO AUTONOMOUS MOBILE ROBOTS

Lesson 6 Learning II Anders Lyhne Christensen, D6.05, INTRODUCTION TO AUTONOMOUS MOBILE ROBOTS Lesson 6 Learning II Anders Lyhne Christensen, D6.05, anders.christensen@iscte.pt INTRODUCTION TO AUTONOMOUS MOBILE ROBOTS First: Quick Background in Neural Nets Some of earliest work in neural networks

More information

Search e Fall /18/15

Search e Fall /18/15 Sample Efficient Policy Click to edit Master title style Search Click to edit Emma Master Brunskill subtitle style 15-889e Fall 2015 11 Sample Efficient RL Objectives Probably Approximately Correct Minimizing

More information

10-1 MMSE Estimation S. Lall, Stanford

10-1 MMSE Estimation S. Lall, Stanford 0 - MMSE Estimation S. Lall, Stanford 20.02.02.0 0 - MMSE Estimation Estimation given a pdf Minimizing the mean square error The minimum mean square error (MMSE) estimator The MMSE and the mean-variance

More information

Rich feature hierarchies for accurate object detection and semantic segmentation

Rich feature hierarchies for accurate object detection and semantic segmentation Rich feature hierarchies for accurate object detection and semantic segmentation Ross Girshick, Jeff Donahue, Trevor Darrell, Jitendra Malik UC Berkeley Tech Report @ http://arxiv.org/abs/1311.2524! Detection

More information

Semi-Supervised Disentangling of Causal Factors. Sargur N. Srihari

Semi-Supervised Disentangling of Causal Factors. Sargur N. Srihari Semi-Supervised Disentangling of Causal Factors Sargur N. srihari@cedar.buffalo.edu 1 Topics in Representation Learning 1. Greedy Layer-Wise Unsupervised Pretraining 2. Transfer Learning and Domain Adaptation

More information

Convolutional and LSTM Neural Networks

Convolutional and LSTM Neural Networks Convolutional and LSTM Neural Networks Vanessa Jurtz January 12, 2016 Contents Neural networks and GPUs Lasagne Peptide binding to MHC class II molecules Convolutional Neural Networks (CNN) Recurrent and

More information

B657: Final Project Report Holistically-Nested Edge Detection

B657: Final Project Report Holistically-Nested Edge Detection B657: Final roject Report Holistically-Nested Edge Detection Mingze Xu & Hanfei Mei May 4, 2016 Abstract Holistically-Nested Edge Detection (HED), which is a novel edge detection method based on fully

More information

CS6501: Deep Learning for Visual Recognition. GenerativeAdversarial Networks (GANs)

CS6501: Deep Learning for Visual Recognition. GenerativeAdversarial Networks (GANs) CS6501: Deep Learning for Visual Recognition GenerativeAdversarial Networks (GANs) Today s Class Adversarial Examples Input Optimization Generative Adversarial Networks (GANs) Conditional GANs Style-Transfer

More information

Skin cancer reorganization and classification with deep neural network

Skin cancer reorganization and classification with deep neural network Skin cancer reorganization and classification with deep neural network Hao Chang 1 1. Department of Genetics, Yale University School of Medicine 2. Email: changhao86@gmail.com Abstract As one kind of skin

More information

A Deep Learning Approach for Subject Independent Emotion Recognition from Facial Expressions

A Deep Learning Approach for Subject Independent Emotion Recognition from Facial Expressions A Deep Learning Approach for Subject Independent Emotion Recognition from Facial Expressions VICTOR-EMIL NEAGOE *, ANDREI-PETRU BĂRAR *, NICU SEBE **, PAUL ROBITU * * Faculty of Electronics, Telecommunications

More information

Classification of breast cancer histology images using transfer learning

Classification of breast cancer histology images using transfer learning Classification of breast cancer histology images using transfer learning Sulaiman Vesal 1 ( ), Nishant Ravikumar 1, AmirAbbas Davari 1, Stephan Ellmann 2, Andreas Maier 1 1 Pattern Recognition Lab, Friedrich-Alexander-Universität

More information

The Importance of Time in Visual Attention Models

The Importance of Time in Visual Attention Models The Importance of Time in Visual Attention Models Degree s Thesis Audiovisual Systems Engineering Author: Advisors: Marta Coll Pol Xavier Giró-i-Nieto and Kevin Mc Guinness Dublin City University (DCU)

More information

Big Image-Omics Data Analytics for Clinical Outcome Prediction

Big Image-Omics Data Analytics for Clinical Outcome Prediction Big Image-Omics Data Analytics for Clinical Outcome Prediction Junzhou Huang, Ph.D. Associate Professor Dept. Computer Science & Engineering University of Texas at Arlington Dept. CSE, UT Arlington Scalable

More information