MEMORY PREFETCHING WITH NEURAL-NETWORKS: CHALLENGES AND INSIGHTS. Leeor Peled, Shie Mannor, Uri Weiser, Yoav Etsion

Size: px
Start display at page:

Download "MEMORY PREFETCHING WITH NEURAL-NETWORKS: CHALLENGES AND INSIGHTS. Leeor Peled, Shie Mannor, Uri Weiser, Yoav Etsion"

Transcription

1 1 MEMORY PREFETCHING WITH NEURAL-NETWORKS: CHALLENGES AND INSIGHTS Leeor Peled, Shie Mannor, Uri Weiser, Yoav Etsion

2 2 Introduction CPU performs speculation on multiple domains Branch prediction, prefetching, cache replacement, disambiguation Most mechanisms try to predict the future based on history of decisions and outcomes We can learn simple recurring patterns, but complex pattern recognition is a challenge Some decision cannot be classified based on simple history alone

3 3 But why can we even make predictions? What makes workloads so predictable? We claim: Locality is a property of program semantics Program + machine attributes can successfully represent a semantic execution context Machine learning can approximate semantic locality

4 4 Next level in memory prefetching Our goal is to approximate Semantic Locality Higher abstraction level: Locality between program objects Accesses are semantically local if they are related through a sequence of actions Dictated by program semantics (e.g.: cur = cur->next) Correlates actions that are consequential, not necessarily consecutive Not just temporal or spatial adjacency

5 5 Example: Memory accesses on MCF x Range B 1,396, ,396, ,396, ,396, ,396, ,396, ,396, ,396, ,396, ,396, ,396, x Range A Memory address space Various interleaved streams, each has it s own traits and semantics for( ; arc < stop_arcs; arc += nr_group ) { if( arc->ident > BASIC ) { red_cost = arc->cost - arc->tail->potential + arc->head->potential; if (bea_is_dual_infeasible(arc, red_cost)) { basket_sz++; perm[basket_sz]->a = arc; perm[basket_sz]->cost = red_cost; perm[basket_sz]->abs_cost = ABS(red_cost); } } } do { while (perm[l]->abs_cost > cut) l++; while (cut > perm[r]->abs_cost) r--; if( l < r ) { xchange = perm[l]; perm[l] = perm[r]; perm[r] = xchange; } if( l <= r ) { l++; r--; } } while (l <= r ); if( min < r ) sort_basket (min, r); if( l < max && l <= B ) sort_basket (l, max);

6 6 Memory accesses on MCF Zoom into the quicksort range The distinctive pattern represents the repeated pivot partitioning and recursion steps. Could this pattern be recognized by NNs?

7 7 Prefetching with Neural-Networks Premise: machine + program state approximate semantic information; this state can be correlated with memory addresses Problem #1: We must find a useful context Too many (/ few) attributes will result in overfit (/ underfit) Each workload may have its own useful subset of attributes Our attempt to solve this with automated feature selection (contextual bandits, ISCA 15) had many shortcomings (size, collisions, adaptability) Problem #2: we need an efficient way of learning Associative: learn global rules from local examples Represent algorithmic complexity

8 8 Neural Network structure We use a NN (3-5 layers of varying sizes) to train the associations (the predictive context and the resulting address) Context: IP, Address Hint, Access type Branch hist Access hist ` ` Address1 Address 2 Additional predictions (confidence, reuse, )

9 9 Can a NN imitate existing prefetchers? As a first step, we trained over sequences representing the patterns targeted by common prefetchers Prefetcher Pattern Additional state Streamer A A+1 A+2 A+3 A+4 A+5 Strider A A+n A+2n A+3n A+4n A+5n SMS A A+n A+m B B+n B+m Markov A B A C A B A B(33%), A C(66%) VLDP, GHB */DC A A+2 A+3 A+5 A+6 A+8 +2,+1,+2,+1,.. GHB PC/* A 1 B 1 A 2 B 2 A 3 B 3 PC A :{A} n, PC B :{B} n Placebo: learn non-trivial functions: sin(x), LFSR, (x+1) 2

10 Simple function: Sin 10

11 11 Convergence: Sin Convergence speed per NN size (sin)

12 Convergence: Prefetchers + Placebo 12 Convergence speed per sequence

13 13 How to imitate a prefetcher? Prefetchers are not mathematical functions Next elements are computed as deltas over previous elements Need to examine how best to learn each function Learning modes include Function learning: x f(x) [described above] Next element: f(x 1 ) f(x 2 ) Next with history: { f(x 1 ), f(x 2 ) } f(x 3 ) Deltas: (f(x 2 ) - f(x 1 )) (f(x 3 ) - f(x 2 ))

14 14 Learning modes: MSE Error rate per trained relation (mode)

15 15 Training on real memory streams Problem: we don t know the correct address to train Unlike the bandits case, we must select a single address (per NN) Train context state from depth n (denoted as S n ), and propagate forward through the network. Select the associated address A i and do back-propagation Selection starts from default depth, and scans for the next miss Can take into account the current network output - address with min square error will train fastest S n S n-1 S 2 S 1 S 0 ` A n A n-1 A d A 0

16 16 NN prefetching: Recap NN Benefits NN can imitate various prefecthers for MCF No need to select attributes. BP will train the weights to activate the most useful inputs. Questions What are the limits on the complexity of patterns we can learn i.e., different workloads How does the prediction accuracy change with NN Width, depth, structure

17 17 Design space and enhancements Network depth and size of hidden layers 3-6 layers, 256 input nodes (context), 64 output (addr/delta), per hidden layer Network connectivity Full, CNN (various reduction ratios), LSTM (dedicated nodes) Number of output prefetches Up to 4, either as independent networks or with shared network Confidence prediction Predicting usefulness internally through the network, or in an external Bloom filter. Association heuristics Minimal MSE, best match or fixed distance (with multiple NN, we can assign different policies) Dynamic selection of max delta and min association depth

18 NN prefetch speedup (SPEC06) 18

19 19 NN prefetch speedup (kernels) linked data structures

20 21 NN prefetching: Challenges Back-propagation is computationally heavy In this work we did a full *magic* back-prop for every memory access But a real back-prop is too slow for that Possible alternatives: Train only a subset of useful predictions Reduced FP precision Quantized / binarized NNs

21 22 NN prefetching: Challenges Technology for an in-core NN is not yet practical Memristors? Off-core accelerator? Convergence time of online learning We can control NN parameters (learning rate, momentum) Unlike tabular updates (bandits), each BP affects all NN weights. May break previous updates Offline learning is possible, but more difficult Offline traces may not represent real-time workloads

22 23 Conclusions Semantic locality traces the origins of temporal and spatial locality to program semantics NNs do a good job at approximating semantic locality NNs can support a generalized prefetcher that targets diverse access patterns, which are today targeted by dedicated prefetchers An NN-based prefetcher is, however, theoretical; existing technology does not support a feasible (in terms of timing, area, and power) in-core prefetcher.

23 Thank You 24

24 25 Quantized Neural Networks Hubara, Courbariaux, Soudry, El-Yaniv, Bengio A hot topic in ML/DL: Train using quantized / discreet weights In the extreme case binary (+/- 1) Greatly simplifies inference No need for MAC (multiply & accumulates), only simple add/sub operations Training accuracy somewhat harmed, but not more than with stochastic gradient descent (SGD) Overall accuracy impact is shown to be small Much easier to implementable and sustain in HW!

SEMANTIC LOCALITY & CONTEXT BASED PREFETCHING. Leeor Peled 1, Shie Mannor 1, Uri Weiser 1, Yoav Etsion 1,2

SEMANTIC LOCALITY & CONTEXT BASED PREFETCHING. Leeor Peled 1, Shie Mannor 1, Uri Weiser 1, Yoav Etsion 1,2 SEMANTIC LOCALITY & CONTEXT BASED PREFETCHING Leeor Peled 1, Shie Mannor 1, Uri Weiser 1, Yoav Etsion 1,2 1 Electrical Engineering and 2 Computer Science Technion Israel Institute of Technology ICRI-CI

More information

Deep Learning and Its Applications

Deep Learning and Its Applications Convolutional Neural Network and Its Application in Image Recognition Oct 28, 2016 Outline 1 A Motivating Example 2 The Convolutional Neural Network (CNN) Model 3 Training the CNN Model 4 Issues and Recent

More information

Profiling the Performance of Binarized Neural Networks. Daniel Lerner, Jared Pierce, Blake Wetherton, Jialiang Zhang

Profiling the Performance of Binarized Neural Networks. Daniel Lerner, Jared Pierce, Blake Wetherton, Jialiang Zhang Profiling the Performance of Binarized Neural Networks Daniel Lerner, Jared Pierce, Blake Wetherton, Jialiang Zhang 1 Outline Project Significance Prior Work Research Objectives Hypotheses Testing Framework

More information

Ensemble methods in machine learning. Example. Neural networks. Neural networks

Ensemble methods in machine learning. Example. Neural networks. Neural networks Ensemble methods in machine learning Bootstrap aggregating (bagging) train an ensemble of models based on randomly resampled versions of the training set, then take a majority vote Example What if you

More information

CNN optimization. Rassadin A

CNN optimization. Rassadin A CNN optimization Rassadin A. 01.2017-02.2017 What to optimize? Training stage time consumption (CPU / GPU) Inference stage time consumption (CPU / GPU) Training stage memory consumption Inference stage

More information

Inthreads: Code Generation and Microarchitecture

Inthreads: Code Generation and Microarchitecture Inthreads: Code Generation and Microarchitecture Alex Gontmakher Gregory Shklover Assaf Schuster Avi Mendelson 1 Outline Inthreads Introduction Inthreads code generation Microarchitecture 2 Motivation

More information

Perceptron: This is convolution!

Perceptron: This is convolution! Perceptron: This is convolution! v v v Shared weights v Filter = local perceptron. Also called kernel. By pooling responses at different locations, we gain robustness to the exact spatial location of image

More information

Memory Bandwidth and Low Precision Computation. CS6787 Lecture 10 Fall 2018

Memory Bandwidth and Low Precision Computation. CS6787 Lecture 10 Fall 2018 Memory Bandwidth and Low Precision Computation CS6787 Lecture 10 Fall 2018 Memory as a Bottleneck So far, we ve just been talking about compute e.g. techniques to decrease the amount of compute by decreasing

More information

CS489/698: Intro to ML

CS489/698: Intro to ML CS489/698: Intro to ML Lecture 14: Training of Deep NNs Instructor: Sun Sun 1 Outline Activation functions Regularization Gradient-based optimization 2 Examples of activation functions 3 5/28/18 Sun Sun

More information

Neural Network Optimization and Tuning / Spring 2018 / Recitation 3

Neural Network Optimization and Tuning / Spring 2018 / Recitation 3 Neural Network Optimization and Tuning 11-785 / Spring 2018 / Recitation 3 1 Logistics You will work through a Jupyter notebook that contains sample and starter code with explanations and comments throughout.

More information

Challenges motivating deep learning. Sargur N. Srihari

Challenges motivating deep learning. Sargur N. Srihari Challenges motivating deep learning Sargur N. srihari@cedar.buffalo.edu 1 Topics In Machine Learning Basics 1. Learning Algorithms 2. Capacity, Overfitting and Underfitting 3. Hyperparameters and Validation

More information

Memory Bandwidth and Low Precision Computation. CS6787 Lecture 9 Fall 2017

Memory Bandwidth and Low Precision Computation. CS6787 Lecture 9 Fall 2017 Memory Bandwidth and Low Precision Computation CS6787 Lecture 9 Fall 2017 Memory as a Bottleneck So far, we ve just been talking about compute e.g. techniques to decrease the amount of compute by decreasing

More information

Deep Learning with Tensorflow AlexNet

Deep Learning with Tensorflow   AlexNet Machine Learning and Computer Vision Group Deep Learning with Tensorflow http://cvml.ist.ac.at/courses/dlwt_w17/ AlexNet Krizhevsky, Alex, Ilya Sutskever, and Geoffrey E. Hinton, "Imagenet classification

More information

Long Short Term Based Memory Hardware Prefetcher

Long Short Term Based Memory Hardware Prefetcher Lehigh University Lehigh Preserve Theses and Dissertations 2017 Long Short Term Based Memory Hardware Prefetcher Yuan Zeng Lehigh University Follow this and additional works at: http://preserve.lehigh.edu/etd

More information

Machine learning for vision. It s the features, stupid! cathedral. high-rise. Winter Roland Memisevic. Lecture 2, January 26, 2016

Machine learning for vision. It s the features, stupid! cathedral. high-rise. Winter Roland Memisevic. Lecture 2, January 26, 2016 Winter 2016 Lecture 2, Januar 26, 2016 f2? cathedral high-rise f1 A common computer vision pipeline before 2012 1. 2. 3. 4. Find interest points. Crop patches around them. Represent each patch with a sparse

More information

Turing Architecture and CUDA 10 New Features. Minseok Lee, Developer Technology Engineer, NVIDIA

Turing Architecture and CUDA 10 New Features. Minseok Lee, Developer Technology Engineer, NVIDIA Turing Architecture and CUDA 10 New Features Minseok Lee, Developer Technology Engineer, NVIDIA Turing Architecture New SM Architecture Multi-Precision Tensor Core RT Core Turing MPS Inference Accelerated,

More information

An Analysis of the Performance Impact of Wrong-Path Memory References on Out-of-Order and Runahead Execution Processors

An Analysis of the Performance Impact of Wrong-Path Memory References on Out-of-Order and Runahead Execution Processors An Analysis of the Performance Impact of Wrong-Path Memory References on Out-of-Order and Runahead Execution Processors Onur Mutlu Hyesoon Kim David N. Armstrong Yale N. Patt High Performance Systems Group

More information

Analyzing Memory Access Patterns and Optimizing Through Spatial Memory Streaming. Ogün HEPER CmpE 511 Computer Architecture December 24th, 2009

Analyzing Memory Access Patterns and Optimizing Through Spatial Memory Streaming. Ogün HEPER CmpE 511 Computer Architecture December 24th, 2009 Analyzing Memory Access Patterns and Optimizing Through Spatial Memory Streaming Ogün HEPER CmpE 511 Computer Architecture December 24th, 2009 Agenda Introduction Memory Hierarchy Design CPU Speed vs.

More information

How Learning Differs from Optimization. Sargur N. Srihari

How Learning Differs from Optimization. Sargur N. Srihari How Learning Differs from Optimization Sargur N. srihari@cedar.buffalo.edu 1 Topics in Optimization Optimization for Training Deep Models: Overview How learning differs from optimization Risk, empirical

More information

Optimization in the Big Data Regime 5: Parallelization? Sham M. Kakade

Optimization in the Big Data Regime 5: Parallelization? Sham M. Kakade Optimization in the Big Data Regime 5: Parallelization? Sham M. Kakade Machine Learning for Big Data CSE547/STAT548 University of Washington S. M. Kakade (UW) Optimization for Big data 1 / 21 Announcements...

More information

Lecture 2 Notes. Outline. Neural Networks. The Big Idea. Architecture. Instructors: Parth Shah, Riju Pahwa

Lecture 2 Notes. Outline. Neural Networks. The Big Idea. Architecture. Instructors: Parth Shah, Riju Pahwa Instructors: Parth Shah, Riju Pahwa Lecture 2 Notes Outline 1. Neural Networks The Big Idea Architecture SGD and Backpropagation 2. Convolutional Neural Networks Intuition Architecture 3. Recurrent Neural

More information

Training Deep Neural Networks (in parallel)

Training Deep Neural Networks (in parallel) Lecture 9: Training Deep Neural Networks (in parallel) Visual Computing Systems How would you describe this professor? Easy? Mean? Boring? Nerdy? Professor classification task Classifies professors as

More information

CSC 578 Neural Networks and Deep Learning

CSC 578 Neural Networks and Deep Learning CSC 578 Neural Networks and Deep Learning Fall 2018/19 7. Recurrent Neural Networks (Some figures adapted from NNDL book) 1 Recurrent Neural Networks 1. Recurrent Neural Networks (RNNs) 2. RNN Training

More information

Accelerating Convolutional Neural Nets. Yunming Zhang

Accelerating Convolutional Neural Nets. Yunming Zhang Accelerating Convolutional Neural Nets Yunming Zhang Focus Convolutional Neural Nets is the state of the art in classifying the images The models take days to train Difficult for the programmers to tune

More information

Neural Networks. CE-725: Statistical Pattern Recognition Sharif University of Technology Spring Soleymani

Neural Networks. CE-725: Statistical Pattern Recognition Sharif University of Technology Spring Soleymani Neural Networks CE-725: Statistical Pattern Recognition Sharif University of Technology Spring 2013 Soleymani Outline Biological and artificial neural networks Feed-forward neural networks Single layer

More information

Pattern Recognition. Kjell Elenius. Speech, Music and Hearing KTH. March 29, 2007 Speech recognition

Pattern Recognition. Kjell Elenius. Speech, Music and Hearing KTH. March 29, 2007 Speech recognition Pattern Recognition Kjell Elenius Speech, Music and Hearing KTH March 29, 2007 Speech recognition 2007 1 Ch 4. Pattern Recognition 1(3) Bayes Decision Theory Minimum-Error-Rate Decision Rules Discriminant

More information

Opening the black box of Deep Neural Networks via Information (Ravid Shwartz-Ziv and Naftali Tishby) An overview by Philip Amortila and Nicolas Gagné

Opening the black box of Deep Neural Networks via Information (Ravid Shwartz-Ziv and Naftali Tishby) An overview by Philip Amortila and Nicolas Gagné Opening the black box of Deep Neural Networks via Information (Ravid Shwartz-Ziv and Naftali Tishby) An overview by Philip Amortila and Nicolas Gagné 1 Problem: The usual "black box" story: "Despite their

More information

Hyperparameter optimization. CS6787 Lecture 6 Fall 2017

Hyperparameter optimization. CS6787 Lecture 6 Fall 2017 Hyperparameter optimization CS6787 Lecture 6 Fall 2017 Review We ve covered many methods Stochastic gradient descent Step size/learning rate, how long to run Mini-batching Batch size Momentum Momentum

More information

Machine Learning 13. week

Machine Learning 13. week Machine Learning 13. week Deep Learning Convolutional Neural Network Recurrent Neural Network 1 Why Deep Learning is so Popular? 1. Increase in the amount of data Thanks to the Internet, huge amount of

More information

In-Place Activated BatchNorm for Memory- Optimized Training of DNNs

In-Place Activated BatchNorm for Memory- Optimized Training of DNNs In-Place Activated BatchNorm for Memory- Optimized Training of DNNs Samuel Rota Bulò, Lorenzo Porzi, Peter Kontschieder Mapillary Research Paper: https://arxiv.org/abs/1712.02616 Code: https://github.com/mapillary/inplace_abn

More information

Fall 2011 Prof. Hyesoon Kim. Thanks to Prof. Loh & Prof. Prvulovic

Fall 2011 Prof. Hyesoon Kim. Thanks to Prof. Loh & Prof. Prvulovic Fall 2011 Prof. Hyesoon Kim Thanks to Prof. Loh & Prof. Prvulovic Reading: Data prefetch mechanisms, Steven P. Vanderwiel, David J. Lilja, ACM Computing Surveys, Vol. 32, Issue 2 (June 2000) If memory

More information

CS 4510/9010 Applied Machine Learning. Neural Nets. Paula Matuszek Fall copyright Paula Matuszek 2016

CS 4510/9010 Applied Machine Learning. Neural Nets. Paula Matuszek Fall copyright Paula Matuszek 2016 CS 4510/9010 Applied Machine Learning 1 Neural Nets Paula Matuszek Fall 2016 Neural Nets, the very short version 2 A neural net consists of layers of nodes, or neurons, each of which has an activation

More information

Cost Functions in Machine Learning

Cost Functions in Machine Learning Cost Functions in Machine Learning Kevin Swingler Motivation Given some data that reflects measurements from the environment We want to build a model that reflects certain statistics about that data Something

More information

Fuzzy Set Theory in Computer Vision: Example 3, Part II

Fuzzy Set Theory in Computer Vision: Example 3, Part II Fuzzy Set Theory in Computer Vision: Example 3, Part II Derek T. Anderson and James M. Keller FUZZ-IEEE, July 2017 Overview Resource; CS231n: Convolutional Neural Networks for Visual Recognition https://github.com/tuanavu/stanford-

More information

COMPUTATIONAL INTELLIGENCE SEW (INTRODUCTION TO MACHINE LEARNING) SS18. Lecture 6: k-nn Cross-validation Regularization

COMPUTATIONAL INTELLIGENCE SEW (INTRODUCTION TO MACHINE LEARNING) SS18. Lecture 6: k-nn Cross-validation Regularization COMPUTATIONAL INTELLIGENCE SEW (INTRODUCTION TO MACHINE LEARNING) SS18 Lecture 6: k-nn Cross-validation Regularization LEARNING METHODS Lazy vs eager learning Eager learning generalizes training data before

More information

Machine Learning Basics: Stochastic Gradient Descent. Sargur N. Srihari

Machine Learning Basics: Stochastic Gradient Descent. Sargur N. Srihari Machine Learning Basics: Stochastic Gradient Descent Sargur N. srihari@cedar.buffalo.edu 1 Topics 1. Learning Algorithms 2. Capacity, Overfitting and Underfitting 3. Hyperparameters and Validation Sets

More information

Comparing Dropout Nets to Sum-Product Networks for Predicting Molecular Activity

Comparing Dropout Nets to Sum-Product Networks for Predicting Molecular Activity 000 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031 032 033 034 035 036 037 038 039 040 041 042 043 044 045 046 047 048 049 050

More information

Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset. By Joa õ Carreira and Andrew Zisserman Presenter: Zhisheng Huang 03/02/2018

Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset. By Joa õ Carreira and Andrew Zisserman Presenter: Zhisheng Huang 03/02/2018 Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset By Joa õ Carreira and Andrew Zisserman Presenter: Zhisheng Huang 03/02/2018 Outline: Introduction Action classification architectures

More information

Text Categorization. Foundations of Statistic Natural Language Processing The MIT Press1999

Text Categorization. Foundations of Statistic Natural Language Processing The MIT Press1999 Text Categorization Foundations of Statistic Natural Language Processing The MIT Press1999 Outline Introduction Decision Trees Maximum Entropy Modeling (optional) Perceptrons K Nearest Neighbor Classification

More information

DECISION TREES & RANDOM FORESTS X CONVOLUTIONAL NEURAL NETWORKS

DECISION TREES & RANDOM FORESTS X CONVOLUTIONAL NEURAL NETWORKS DECISION TREES & RANDOM FORESTS X CONVOLUTIONAL NEURAL NETWORKS Deep Neural Decision Forests Microsoft Research Cambridge UK, ICCV 2015 Decision Forests, Convolutional Networks and the Models in-between

More information

CS273 Midterm Exam Introduction to Machine Learning: Winter 2015 Tuesday February 10th, 2014

CS273 Midterm Exam Introduction to Machine Learning: Winter 2015 Tuesday February 10th, 2014 CS273 Midterm Eam Introduction to Machine Learning: Winter 2015 Tuesday February 10th, 2014 Your name: Your UCINetID (e.g., myname@uci.edu): Your seat (row and number): Total time is 80 minutes. READ THE

More information

COMP9444 Neural Networks and Deep Learning 5. Geometry of Hidden Units

COMP9444 Neural Networks and Deep Learning 5. Geometry of Hidden Units COMP9 8s Geometry of Hidden Units COMP9 Neural Networks and Deep Learning 5. Geometry of Hidden Units Outline Geometry of Hidden Unit Activations Limitations of -layer networks Alternative transfer functions

More information

Tutorial on Keras CAP ADVANCED COMPUTER VISION SPRING 2018 KISHAN S ATHREY

Tutorial on Keras CAP ADVANCED COMPUTER VISION SPRING 2018 KISHAN S ATHREY Tutorial on Keras CAP 6412 - ADVANCED COMPUTER VISION SPRING 2018 KISHAN S ATHREY Deep learning packages TensorFlow Google PyTorch Facebook AI research Keras Francois Chollet (now at Google) Chainer Company

More information

COMP 551 Applied Machine Learning Lecture 16: Deep Learning

COMP 551 Applied Machine Learning Lecture 16: Deep Learning COMP 551 Applied Machine Learning Lecture 16: Deep Learning Instructor: Ryan Lowe (ryan.lowe@cs.mcgill.ca) Slides mostly by: Class web page: www.cs.mcgill.ca/~hvanho2/comp551 Unless otherwise noted, all

More information

CS 4510/9010 Applied Machine Learning

CS 4510/9010 Applied Machine Learning CS 4510/9010 Applied Machine Learning Neural Nets Paula Matuszek Spring, 2015 1 Neural Nets, the very short version A neural net consists of layers of nodes, or neurons, each of which has an activation

More information

4.12 Generalization. In back-propagation learning, as many training examples as possible are typically used.

4.12 Generalization. In back-propagation learning, as many training examples as possible are typically used. 1 4.12 Generalization In back-propagation learning, as many training examples as possible are typically used. It is hoped that the network so designed generalizes well. A network generalizes well when

More information

17/05/2018. Outline. Outline. Divide and Conquer. Control Abstraction for Divide &Conquer. Outline. Module 2: Divide and Conquer

17/05/2018. Outline. Outline. Divide and Conquer. Control Abstraction for Divide &Conquer. Outline. Module 2: Divide and Conquer Module 2: Divide and Conquer Divide and Conquer Control Abstraction for Divide &Conquer 1 Recurrence equation for Divide and Conquer: If the size of problem p is n and the sizes of the k sub problems are

More information

Index. Springer Nature Switzerland AG 2019 B. Moons et al., Embedded Deep Learning,

Index. Springer Nature Switzerland AG 2019 B. Moons et al., Embedded Deep Learning, Index A Algorithmic noise tolerance (ANT), 93 94 Application specific instruction set processors (ASIPs), 115 116 Approximate computing application level, 95 circuits-levels, 93 94 DAS and DVAS, 107 110

More information

Multimedia Systems Video II (Video Coding) Mahdi Amiri April 2012 Sharif University of Technology

Multimedia Systems Video II (Video Coding) Mahdi Amiri April 2012 Sharif University of Technology Course Presentation Multimedia Systems Video II (Video Coding) Mahdi Amiri April 2012 Sharif University of Technology Video Coding Correlation in Video Sequence Spatial correlation Similar pixels seem

More information

Lecture 19 Sorting Goodrich, Tamassia

Lecture 19 Sorting Goodrich, Tamassia Lecture 19 Sorting 7 2 9 4 2 4 7 9 7 2 2 7 9 4 4 9 7 7 2 2 9 9 4 4 2004 Goodrich, Tamassia Outline Review 3 simple sorting algorithms: 1. selection Sort (in previous course) 2. insertion Sort (in previous

More information

A Lightweight YOLOv2:

A Lightweight YOLOv2: FPGA2018 @Monterey A Lightweight YOLOv2: A Binarized CNN with a Parallel Support Vector Regression for an FPGA Hiroki Nakahara, Haruyoshi Yonekawa, Tomoya Fujii, Shimpei Sato Tokyo Institute of Technology,

More information

Simple Model Selection Cross Validation Regularization Neural Networks

Simple Model Selection Cross Validation Regularization Neural Networks Neural Nets: Many possible refs e.g., Mitchell Chapter 4 Simple Model Selection Cross Validation Regularization Neural Networks Machine Learning 10701/15781 Carlos Guestrin Carnegie Mellon University February

More information

CMU Lecture 18: Deep learning and Vision: Convolutional neural networks. Teacher: Gianni A. Di Caro

CMU Lecture 18: Deep learning and Vision: Convolutional neural networks. Teacher: Gianni A. Di Caro CMU 15-781 Lecture 18: Deep learning and Vision: Convolutional neural networks Teacher: Gianni A. Di Caro DEEP, SHALLOW, CONNECTED, SPARSE? Fully connected multi-layer feed-forward perceptrons: More powerful

More information

Pouya Kousha Fall 2018 CSE 5194 Prof. DK Panda

Pouya Kousha Fall 2018 CSE 5194 Prof. DK Panda Pouya Kousha Fall 2018 CSE 5194 Prof. DK Panda 1 Observe novel applicability of DL techniques in Big Data Analytics. Applications of DL techniques for common Big Data Analytics problems. Semantic indexing

More information

Machine Learning (CSE 446): Concepts & the i.i.d. Supervised Learning Paradigm

Machine Learning (CSE 446): Concepts & the i.i.d. Supervised Learning Paradigm Machine Learning (CSE 446): Concepts & the i.i.d. Supervised Learning Paradigm Sham M Kakade c 2018 University of Washington cse446-staff@cs.washington.edu 1 / 17 Review 1 / 17 Decision Tree: Making a

More information

Table of Contents. What Really is a Hidden Unit? Visualizing Feed-Forward NNs. Visualizing Convolutional NNs. Visualizing Recurrent NNs

Table of Contents. What Really is a Hidden Unit? Visualizing Feed-Forward NNs. Visualizing Convolutional NNs. Visualizing Recurrent NNs Table of Contents What Really is a Hidden Unit? Visualizing Feed-Forward NNs Visualizing Convolutional NNs Visualizing Recurrent NNs Visualizing Attention Visualizing High Dimensional Data What do visualizations

More information

Deep Learning & Neural Networks

Deep Learning & Neural Networks Deep Learning & Neural Networks Machine Learning CSE4546 Sham Kakade University of Washington November 29, 2016 Sham Kakade 1 Announcements: HW4 posted Poster Session Thurs, Dec 8 Today: Review: EM Neural

More information

Keras: Handwritten Digit Recognition using MNIST Dataset

Keras: Handwritten Digit Recognition using MNIST Dataset Keras: Handwritten Digit Recognition using MNIST Dataset IIT PATNA January 31, 2018 1 / 30 OUTLINE 1 Keras: Introduction 2 Installing Keras 3 Keras: Building, Testing, Improving A Simple Network 2 / 30

More information

Why DNN Works for Speech and How to Make it More Efficient?

Why DNN Works for Speech and How to Make it More Efficient? Why DNN Works for Speech and How to Make it More Efficient? Hui Jiang Department of Electrical Engineering and Computer Science Lassonde School of Engineering, York University, CANADA Joint work with Y.

More information

Graph Neural Network. learning algorithm and applications. Shujia Zhang

Graph Neural Network. learning algorithm and applications. Shujia Zhang Graph Neural Network learning algorithm and applications Shujia Zhang What is Deep Learning? Enable computers to mimic human behaviour Artificial Intelligence Machine Learning Subset of ML algorithms using

More information

The OpenVX Computer Vision and Neural Network Inference

The OpenVX Computer Vision and Neural Network Inference The OpenVX Computer and Neural Network Inference Standard for Portable, Efficient Code Radhakrishna Giduthuri Editor, OpenVX Khronos Group radha.giduthuri@amd.com @RadhaGiduthuri Copyright 2018 Khronos

More information

Business Club. Decision Trees

Business Club. Decision Trees Business Club Decision Trees Business Club Analytics Team December 2017 Index 1. Motivation- A Case Study 2. The Trees a. What is a decision tree b. Representation 3. Regression v/s Classification 4. Building

More information

Wenisch Final Review. Fall 2007 Prof. Thomas Wenisch EECS 470. Slide 1

Wenisch Final Review. Fall 2007 Prof. Thomas Wenisch  EECS 470. Slide 1 Final Review Fall 2007 Prof. Thomas Wenisch http://www.eecs.umich.edu/courses/eecs470 Slide 1 Announcements Wenisch 2007 Exam is Monday, 12/17 4 6 in this room I recommend bringing a scientific calculator

More information

Neural Network Learning. Today s Lecture. Continuation of Neural Networks. Artificial Neural Networks. Lecture 24: Learning 3. Victor R.

Neural Network Learning. Today s Lecture. Continuation of Neural Networks. Artificial Neural Networks. Lecture 24: Learning 3. Victor R. Lecture 24: Learning 3 Victor R. Lesser CMPSCI 683 Fall 2010 Today s Lecture Continuation of Neural Networks Artificial Neural Networks Compose of nodes/units connected by links Each link has a numeric

More information

COMP9444 Neural Networks and Deep Learning 7. Image Processing. COMP9444 c Alan Blair, 2017

COMP9444 Neural Networks and Deep Learning 7. Image Processing. COMP9444 c Alan Blair, 2017 COMP9444 Neural Networks and Deep Learning 7. Image Processing COMP9444 17s2 Image Processing 1 Outline Image Datasets and Tasks Convolution in Detail AlexNet Weight Initialization Batch Normalization

More information

ECE7995 (3) Basis of Caching and Prefetching --- Locality

ECE7995 (3) Basis of Caching and Prefetching --- Locality ECE7995 (3) Basis of Caching and Prefetching --- Locality 1 What s locality? Temporal locality is a property inherent to programs and independent of their execution environment. Temporal locality: the

More information

ECG782: Multidimensional Digital Signal Processing

ECG782: Multidimensional Digital Signal Processing ECG782: Multidimensional Digital Signal Processing Object Recognition http://www.ee.unlv.edu/~b1morris/ecg782/ 2 Outline Knowledge Representation Statistical Pattern Recognition Neural Networks Boosting

More information

Partitioning Data. IRDS: Evaluation, Debugging, and Diagnostics. Cross-Validation. Cross-Validation for parameter tuning

Partitioning Data. IRDS: Evaluation, Debugging, and Diagnostics. Cross-Validation. Cross-Validation for parameter tuning Partitioning Data IRDS: Evaluation, Debugging, and Diagnostics Charles Sutton University of Edinburgh Training Validation Test Training : Running learning algorithms Validation : Tuning parameters of learning

More information

291 Programming Assignment #3

291 Programming Assignment #3 000 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031 032 033 034 035 036 037 038 039 040 041 042 043 044 045 046 047 048 049 050

More information

Accelerating Binarized Convolutional Neural Networks with Software-Programmable FPGAs

Accelerating Binarized Convolutional Neural Networks with Software-Programmable FPGAs Accelerating Binarized Convolutional Neural Networks with Software-Programmable FPGAs Ritchie Zhao 1, Weinan Song 2, Wentao Zhang 2, Tianwei Xing 3, Jeng-Hau Lin 4, Mani Srivastava 3, Rajesh Gupta 4, Zhiru

More information

Neural Network Neurons

Neural Network Neurons Neural Networks Neural Network Neurons 1 Receives n inputs (plus a bias term) Multiplies each input by its weight Applies activation function to the sum of results Outputs result Activation Functions Given

More information

Parallel Deep Network Training

Parallel Deep Network Training Lecture 19: Parallel Deep Network Training Parallel Computer Architecture and Programming How would you describe this professor? Easy? Mean? Boring? Nerdy? Professor classification task Classifies professors

More information

Protecting Web Servers from Web Robot Traffic

Protecting Web Servers from Web Robot Traffic Wright State University CORE Scholar Kno.e.sis Publications The Ohio Center of Excellence in Knowledge- Enabled Computing (Kno.e.sis) 11-14-2014 Protecting Web Servers from Web Robot Traffic Derek Doran

More information

10-701/15-781, Fall 2006, Final

10-701/15-781, Fall 2006, Final -7/-78, Fall 6, Final Dec, :pm-8:pm There are 9 questions in this exam ( pages including this cover sheet). If you need more room to work out your answer to a question, use the back of the page and clearly

More information

Data Mining. Part 2. Data Understanding and Preparation. 2.4 Data Transformation. Spring Instructor: Dr. Masoud Yaghini. Data Transformation

Data Mining. Part 2. Data Understanding and Preparation. 2.4 Data Transformation. Spring Instructor: Dr. Masoud Yaghini. Data Transformation Data Mining Part 2. Data Understanding and Preparation 2.4 Spring 2010 Instructor: Dr. Masoud Yaghini Outline Introduction Normalization Attribute Construction Aggregation Attribute Subset Selection Discretization

More information

DEEP LEARNING REVIEW. Yann LeCun, Yoshua Bengio & Geoffrey Hinton Nature Presented by Divya Chitimalla

DEEP LEARNING REVIEW. Yann LeCun, Yoshua Bengio & Geoffrey Hinton Nature Presented by Divya Chitimalla DEEP LEARNING REVIEW Yann LeCun, Yoshua Bengio & Geoffrey Hinton Nature 2015 -Presented by Divya Chitimalla What is deep learning Deep learning allows computational models that are composed of multiple

More information

Natural Language Processing

Natural Language Processing Natural Language Processing Classification III Dan Klein UC Berkeley 1 Classification 2 Linear Models: Perceptron The perceptron algorithm Iteratively processes the training set, reacting to training errors

More information

LSTM: An Image Classification Model Based on Fashion-MNIST Dataset

LSTM: An Image Classification Model Based on Fashion-MNIST Dataset LSTM: An Image Classification Model Based on Fashion-MNIST Dataset Kexin Zhang, Research School of Computer Science, Australian National University Kexin Zhang, U6342657@anu.edu.au Abstract. The application

More information

MIT 801. Machine Learning I. [Presented by Anna Bosman] 16 February 2018

MIT 801. Machine Learning I. [Presented by Anna Bosman] 16 February 2018 MIT 801 [Presented by Anna Bosman] 16 February 2018 Machine Learning What is machine learning? Artificial Intelligence? Yes as we know it. What is intelligence? The ability to acquire and apply knowledge

More information

Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling

Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling Authors: Junyoung Chung, Caglar Gulcehre, KyungHyun Cho and Yoshua Bengio Presenter: Yu-Wei Lin Background: Recurrent Neural

More information

PERFORMANCE OF GRID COMPUTING FOR DISTRIBUTED NEURAL NETWORK. Submitted By:Mohnish Malviya & Suny Shekher Pankaj [CSE,7 TH SEM]

PERFORMANCE OF GRID COMPUTING FOR DISTRIBUTED NEURAL NETWORK. Submitted By:Mohnish Malviya & Suny Shekher Pankaj [CSE,7 TH SEM] PERFORMANCE OF GRID COMPUTING FOR DISTRIBUTED NEURAL NETWORK Submitted By:Mohnish Malviya & Suny Shekher Pankaj [CSE,7 TH SEM] All Saints` College Of Technology, Gandhi Nagar, Bhopal. Abstract: In this

More information

Dropout. Sargur N. Srihari This is part of lecture slides on Deep Learning:

Dropout. Sargur N. Srihari This is part of lecture slides on Deep Learning: Dropout Sargur N. srihari@buffalo.edu This is part of lecture slides on Deep Learning: http://www.cedar.buffalo.edu/~srihari/cse676 1 Regularization Strategies 1. Parameter Norm Penalties 2. Norm Penalties

More information

Algorithm Efficiency & Sorting. Algorithm efficiency Big-O notation Searching algorithms Sorting algorithms

Algorithm Efficiency & Sorting. Algorithm efficiency Big-O notation Searching algorithms Sorting algorithms Algorithm Efficiency & Sorting Algorithm efficiency Big-O notation Searching algorithms Sorting algorithms Overview Writing programs to solve problem consists of a large number of decisions how to represent

More information

Machine Learning. Deep Learning. Eric Xing (and Pengtao Xie) , Fall Lecture 8, October 6, Eric CMU,

Machine Learning. Deep Learning. Eric Xing (and Pengtao Xie) , Fall Lecture 8, October 6, Eric CMU, Machine Learning 10-701, Fall 2015 Deep Learning Eric Xing (and Pengtao Xie) Lecture 8, October 6, 2015 Eric Xing @ CMU, 2015 1 A perennial challenge in computer vision: feature engineering SIFT Spin image

More information

27: Hybrid Graphical Models and Neural Networks

27: Hybrid Graphical Models and Neural Networks 10-708: Probabilistic Graphical Models 10-708 Spring 2016 27: Hybrid Graphical Models and Neural Networks Lecturer: Matt Gormley Scribes: Jakob Bauer Otilia Stretcu Rohan Varma 1 Motivation We first look

More information

15-740/ Computer Architecture

15-740/ Computer Architecture 15-740/18-740 Computer Architecture Lecture 16: Runahead and OoO Wrap-Up Prof. Onur Mutlu Carnegie Mellon University Fall 2011, 10/17/2011 Review Set 9 Due this Wednesday (October 19) Wilkes, Slave Memories

More information

CUDA Optimizations WS Intelligent Robotics Seminar. Universität Hamburg WS Intelligent Robotics Seminar Praveen Kulkarni

CUDA Optimizations WS Intelligent Robotics Seminar. Universität Hamburg WS Intelligent Robotics Seminar Praveen Kulkarni CUDA Optimizations WS 2014-15 Intelligent Robotics Seminar 1 Table of content 1 Background information 2 Optimizations 3 Summary 2 Table of content 1 Background information 2 Optimizations 3 Summary 3

More information

11. Neural Network Regularization

11. Neural Network Regularization 11. Neural Network Regularization CS 519 Deep Learning, Winter 2016 Fuxin Li With materials from Andrej Karpathy, Zsolt Kira Preventing overfitting Approach 1: Get more data! Always best if possible! If

More information

COMP 551 Applied Machine Learning Lecture 14: Neural Networks

COMP 551 Applied Machine Learning Lecture 14: Neural Networks COMP 551 Applied Machine Learning Lecture 14: Neural Networks Instructor: (jpineau@cs.mcgill.ca) Class web page: www.cs.mcgill.ca/~jpineau/comp551 Unless otherwise noted, all material posted for this course

More information

Radial Basis Function Neural Network Classifier

Radial Basis Function Neural Network Classifier Recognition of Unconstrained Handwritten Numerals by a Radial Basis Function Neural Network Classifier Hwang, Young-Sup and Bang, Sung-Yang Department of Computer Science & Engineering Pohang University

More information

TDT Coarse-Grained Multithreading. Review on ILP. Multi-threaded execution. Contents. Fine-Grained Multithreading

TDT Coarse-Grained Multithreading. Review on ILP. Multi-threaded execution. Contents. Fine-Grained Multithreading Review on ILP TDT 4260 Chap 5 TLP & Hierarchy What is ILP? Let the compiler find the ILP Advantages? Disadvantages? Let the HW find the ILP Advantages? Disadvantages? Contents Multi-threading Chap 3.5

More information

Parallelism. CS6787 Lecture 8 Fall 2017

Parallelism. CS6787 Lecture 8 Fall 2017 Parallelism CS6787 Lecture 8 Fall 2017 So far We ve been talking about algorithms We ve been talking about ways to optimize their parameters But we haven t talked about the underlying hardware How does

More information

Deep Learning. Volker Tresp Summer 2014

Deep Learning. Volker Tresp Summer 2014 Deep Learning Volker Tresp Summer 2014 1 Neural Network Winter and Revival While Machine Learning was flourishing, there was a Neural Network winter (late 1990 s until late 2000 s) Around 2010 there

More information

Accelerated tokamak transport simulations

Accelerated tokamak transport simulations Accelerated tokamak transport simulations via Neural-Network based regression of TGLF turbulent energy, particle and momentum fluxes by Teobaldo Luda 1 O. Meneghini 2, S. Smith 2, G. Staebler 2 J. Candy

More information

Deep Learning. Practical introduction with Keras JORDI TORRES 27/05/2018. Chapter 3 JORDI TORRES

Deep Learning. Practical introduction with Keras JORDI TORRES 27/05/2018. Chapter 3 JORDI TORRES Deep Learning Practical introduction with Keras Chapter 3 27/05/2018 Neuron A neural network is formed by neurons connected to each other; in turn, each connection of one neural network is associated

More information

CS229 Lecture notes. Raphael John Lamarre Townshend

CS229 Lecture notes. Raphael John Lamarre Townshend CS229 Lecture notes Raphael John Lamarre Townshend Decision Trees We now turn our attention to decision trees, a simple yet flexible class of algorithms. We will first consider the non-linear, region-based

More information

Nearest Neighbor Predictors

Nearest Neighbor Predictors Nearest Neighbor Predictors September 2, 2018 Perhaps the simplest machine learning prediction method, from a conceptual point of view, and perhaps also the most unusual, is the nearest-neighbor method,

More information

Lecture 2: Pipelining Basics. Today: chapter 1 wrap-up, basic pipelining implementation (Sections A.1 - A.4)

Lecture 2: Pipelining Basics. Today: chapter 1 wrap-up, basic pipelining implementation (Sections A.1 - A.4) Lecture 2: Pipelining Basics Today: chapter 1 wrap-up, basic pipelining implementation (Sections A.1 - A.4) 1 Defining Fault, Error, and Failure A fault produces a latent error; it becomes effective when

More information

Conflict Graphs for Parallel Stochastic Gradient Descent

Conflict Graphs for Parallel Stochastic Gradient Descent Conflict Graphs for Parallel Stochastic Gradient Descent Darshan Thaker*, Guneet Singh Dhillon* Abstract We present various methods for inducing a conflict graph in order to effectively parallelize Pegasos.

More information

Analysis of Algorithms. Unit 4 - Analysis of well known Algorithms

Analysis of Algorithms. Unit 4 - Analysis of well known Algorithms Analysis of Algorithms Unit 4 - Analysis of well known Algorithms 1 Analysis of well known Algorithms Brute Force Algorithms Greedy Algorithms Divide and Conquer Algorithms Decrease and Conquer Algorithms

More information