Introduction to Neural Networks Gianluca Pollastri, Head of Lab School of Computer Science and Informatics and Complex and Adaptive Systems Labs University.

Slides:



Advertisements
Similar presentations
Artificial Neural Networks
Advertisements

Bioinspired Computing Lecture 16
Pattern Association.
A Brief Overview of Neural Networks By Rohit Dua, Samuel A. Mulder, Steve E. Watkins, and Donald C. Wunsch.
Learning in Neural and Belief Networks - Feed Forward Neural Network 2001 년 3 월 28 일 안순길.
1 Machine Learning: Lecture 4 Artificial Neural Networks (Based on Chapter 4 of Mitchell T.., Machine Learning, 1997)
CSC321: 2011 Introduction to Neural Networks and Machine Learning Lecture 7: Learning in recurrent networks Geoffrey Hinton.
Tuomas Sandholm Carnegie Mellon University Computer Science Department
Kostas Kontogiannis E&CE
Machine Learning: Connectionist McCulloch-Pitts Neuron Perceptrons Multilayer Networks Support Vector Machines Feedback Networks Hopfield Networks.
Machine Learning Neural Networks
Non-linear classification problem using NN Fainan May 2006 Pattern Classification and Machine Learning Course Three layers Feedforward Neural Network (FFNN)
Neural NetworksNN 11 Neural Networks Teacher: Elena Marchiori R4.47 Assistant: Kees Jong S2.22
Branch Prediction with Neural- Networks: Hidden Layers and Recurrent Connections Andrew Smith CSE Dept. June 10, 2004.
Neural Networks Basic concepts ArchitectureOperation.
I welcome you all to this presentation On: Neural Network Applications Systems Engineering Dept. KFUPM Imran Nadeem & Naveed R. Butt &
1 Chapter 11 Neural Networks. 2 Chapter 11 Contents (1) l Biological Neurons l Artificial Neurons l Perceptrons l Multilayer Neural Networks l Backpropagation.
Rutgers CS440, Fall 2003 Neural networks Reading: Ch. 20, Sec. 5, AIMA 2 nd Ed.
Neural Networks Marco Loog.
Introduction to Recurrent neural networks (RNN), Long short-term memory (LSTM) Wenjie Pei In this coffee talk, I would like to present you some basic.
Neural Networks An Introduction.
SOMTIME: AN ARTIFICIAL NEURAL NETWORK FOR TOPOLOGICAL AND TEMPORAL CORRELATION FOR SPATIOTEMPORAL PATTERN LEARNING.
Soft Computing Colloquium 2 Selection of neural network, Hybrid neural networks.
Neural Networks. Plan Perceptron  Linear discriminant Associative memories  Hopfield networks  Chaotic networks Multilayer perceptron  Backpropagation.
Presentation on Neural Networks.. Basics Of Neural Networks Neural networks refers to a connectionist model that simulates the biophysical information.
Computer Science and Engineering
Multi Layer NN and Bit-True Modeling of These Networks SILab presentation Ali Ahmadi September 2007.
IE 585 Introduction to Neural Networks. 2 Modeling Continuum Unarticulated Wisdom Articulated Qualitative Models Theoretic (First Principles) Models Empirical.
Artificial Neural Networks An Overview and Analysis.
Neural Networks Ellen Walker Hiram College. Connectionist Architectures Characterized by (Rich & Knight) –Large number of very simple neuron-like processing.
© Copyright 2004 ECE, UM-Rolla. All rights reserved A Brief Overview of Neural Networks By Rohit Dua, Samuel A. Mulder, Steve E. Watkins, and Donald C.
Self organizing maps 1 iCSC2014, Juan López González, University of Oviedo Self organizing maps A visualization technique with data dimension reduction.
1 Chapter 6: Artificial Neural Networks Part 2 of 3 (Sections 6.4 – 6.6) Asst. Prof. Dr. Sukanya Pongsuparb Dr. Srisupa Palakvangsa Na Ayudhya Dr. Benjarath.
Appendix B: An Example of Back-propagation algorithm
1 Chapter 11 Neural Networks. 2 Chapter 11 Contents (1) l Biological Neurons l Artificial Neurons l Perceptrons l Multilayer Neural Networks l Backpropagation.
CS344: Introduction to Artificial Intelligence (associated lab: CS386) Pushpak Bhattacharyya CSE Dept., IIT Bombay Lecture 31: Feedforward N/W; sigmoid.
Multi-Layer Perceptron
ICS 586: Neural Networks Dr. Lahouari Ghouti Information & Computer Science Department.
CS621 : Artificial Intelligence
ECE 6504: Deep Learning for Perception Dhruv Batra Virginia Tech Topics: –Recurrent Neural Networks (RNNs) –BackProp Through Time (BPTT) –Vanishing / Exploding.
Neural Networks Teacher: Elena Marchiori R4.47 Assistant: Kees Jong S2.22
Introduction to Neural Networks Jianfeng Feng School of Cognitive and Computing Sciences Spring 2001.
Supervised Sequence Labelling with Recurrent Neural Networks PRESENTED BY: KUNAL PARMAR UHID:
COSC 4426 AJ Boulay Julia Johnson Artificial Neural Networks: Introduction to Soft Computing (Textbook)
Introduction to Neural Networks Freek Stulp. 2 Overview Biological Background Artificial Neuron Classes of Neural Networks 1. Perceptrons 2. Multi-Layered.
Artificial Neural Networks (ANN). Artificial Neural Networks First proposed in 1940s as an attempt to simulate the human brain’s cognitive learning processes.
Object Recognizing. Deep Learning Success in 2012 DeepNet and speech processing.
Neural Networks Lecture 11: Learning in recurrent networks Geoffrey Hinton.
Neural Networks Lecture 4 out of 4. Practical Considerations Input Architecture Output.
語音訊號處理之初步實驗 NTU Speech Lab 指導教授: 李琳山 助教: 熊信寬
Machine Learning Artificial Neural Networks MPλ ∀ Stergiou Theodoros 1.
1 Introduction to Neural Networks Recurrent Neural Networks.
1 Neural Networks MUMT 611 Philippe Zaborowski April 2005.
Neural Network Architecture Session 2
Matt Gormley Lecture 16 October 24, 2016
CSSE463: Image Recognition Day 17
A critical review of RNN for sequence learning Zachary C
Convolutional Neural Networks
Artificial Neural Networks
CSSE463: Image Recognition Day 17
Capabilities of Threshold Neurons
CSSE463: Image Recognition Day 17
CSSE463: Image Recognition Day 13
A Dynamic System Analysis of Simultaneous Recurrent Neural Network
CSSE463: Image Recognition Day 17
CSSE463: Image Recognition Day 17
Deep Learning Authors: Yann LeCun, Yoshua Bengio, Geoffrey Hinton
CSC321: Neural Networks Lecture 11: Learning in recurrent networks
August 8, 2006 Danny Budik, Itamar Elhanany Machine Intelligence Lab
CS621: Artificial Intelligence Lecture 18: Feedforward network contd
Presentation transcript:

Introduction to Neural Networks Gianluca Pollastri, Head of Lab School of Computer Science and Informatics and Complex and Adaptive Systems Labs University College Dublin

Credits Geoffrey Hinton, University of Toronto.  borrowed some of his slides for “Neural Networks” and “Computation in Neural Networks” courses. Paolo Frasconi, University of Florence.  This guy taught me Neural Networks in the first place (*and* I borrowed some of his slides too!).

Recurrent Neural Networks (RNN) One of the earliest versions: Jeffrey Elman, 1990, Cognitive Science. Problem: it isn’t easy to represent time with Feedforward Neural Nets: usually time is represented with space. Attempt to design networks with memory.

RNNs The idea is having discrete time steps, and considering the hidden layer at time t-1 as an input at time t. This effectively removes cycles: we can model the network using an FFNN, and model memory explicitly.

ItIt XtXt OtOt d d = delay element

BPTT BackPropagation Through Time. If O t is the output at time t, I t the input at time t, and X t the memory (hidden) at time t, we can model the dependencies as follows:

BPTT We can model both f() and g() with (possibly multilayered) networks. We can transform the recurrent network by unrolling it in time. Backpropagation works on any DAG. An RNN becomes one once it’s unrolled.

ItIt XtXt OtOt d d = delay element

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

gradient in BPTT GRADIENT(I,O,T) { # I=inputs, O=outputs, T=targets T := size(O); X 0 := 0; for t := 1..T X t := f( X t-1, I t ); for t := 1..T { O t := g( X t, I t ); g.gradient( O t - T t ); δ t = g.deltas( O t - T t ); } for t := T..1 f.gradient( δ t ); δ t-1 += f.deltas( δ t ); }

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

ItIt XtXt OtOt I t+1 X t+1 O t+1 I t-1 X t-1 O t-1 I t+2 X t+2 O t+2 I t-2 X t-2 O t-2

What I will talk about Neurons Multi-Layered Neural Networks:  Basic learning algorithm  Expressive power  Classification How can we *actually* train Neural Networks:  Speeding up training  Learning just right (not too little, not too much)  Figuring out you got it right Feed-back networks?  Anecdotes on real feed-back networks (Hopfield Nets, Boltzmann Machines)  Recurrent Neural Networks  Bidirectional RNN  2D-RNN Concluding remarks

Bidirectional Recurrent Neural Networks (BRNN)

BRNN F t =  ( F t-1, U t ) B t =  ( B t+1, U t ) Y t =  ( F t, B t, U t )  ()  () ed  () are realised with NN  (),  () and  () are independent from t: stationary

BRNN F t =  ( F t-1, U t ) B t =  ( B t+1, U t ) Y t =  ( F t, B t, U t )  ()  () ed  () are realised with NN  (),  () and  () are independent from t: stationary

BRNN F t =  ( F t-1, U t ) B t =  ( B t+1, U t ) Y t =  ( F t, B t, U t )  ()  () ed  () are realised with NN  (),  () and  () are independent from t: stationary

BRNN F t =  ( F t-1, U t ) B t =  ( B t+1, U t ) Y t =  ( F t, B t, U t )  ()  () ed  () are realised with NN  (),  () and  () are independent from t: stationary

Inference in BRNNs FORWARD(U) { T  size(U); F 0  B T+1  0; for t  1..T F t =  ( F t-1, U t ); for t  T..1 B t =  ( B t+1, U t ); for t  1..T Y t =  ( F t, B t, U t ); return Y; }

Learning in BRNNs GRADIENT(U,Y) { T  size(U); F 0  B T+1  0; for t  1..T F t =  ( F t-1, U t ); for t  T..1 B t =  ( B t+1, U t ); for t  1..T { Y t =  ( F t, B t, U t ); [δ F t, δ B t ] = .backprop&gradient( Y t - Y t ); } for t  T..1 δ F t-1 += .backprop&gradient(δ F t ); for t  1..T δ B t+1 += .backprop&gradient(δ B t ); }

What I will talk about Neurons Multi-Layered Neural Networks:  Basic learning algorithm  Expressive power  Classification How can we *actually* train Neural Networks:  Speeding up training  Learning just right (not too little, not too much)  Figuring out you got it right Feed-back networks?  Anecdotes on real feed-back networks (Hopfield Nets, Boltzmann Machines)  Recurrent Neural Networks  Bidirectional RNN  2D-RNN Concluding remarks

2D RNNs Pollastri & Baldi 2002, Bioinformatics Baldi & Pollastri 2003, JMLR

2D RNNs