Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

H2O Open Source Deep Learning, Arno Candel 03-20-14

14,968 views

Published on

More information in our Deep Learning webinar: http://www.slideshare.net/0xdata/h2-o-deeplearningarnocandel052114

Latest slide deck: http://www.slideshare.net/0xdata/h2o-distributed-deep-learning-by-arno-candel-071614

- Powered by the open source machine learning software H2O.ai. Contributors welcome at: https://github.com/h2oai
- To view videos on H2O open source machine learning software, go to: https://www.youtube.com/user/0xdata

Published in: Technology

H2O Open Source Deep Learning, Arno Candel 03-20-14

  1. 1. Deep Learning with H2O ! H2O.ai
 Scalable In-Memory Machine Learning ! H20 Meetup, Mountain View, 3/20/14 Arno Candel
  2. 2. Who am I? PhD in Computational Physics, 2005
 from ETH Zurich Switzerland ! 6 years at SLAC - Accelerator Physics Modeling 2 years at Skytree, Inc - Machine Learning 3 months at 0xdata/H2O - Machine Learning ! 10+ years in HPC, C++, MPI, Supercomputing Arno Candel
  3. 3. Outline Intro Theory Implementation Results MNIST handwritten digits classification Live Demo Prostate cancer classification and age regression text classification
  4. 4. Distributed in-memory math platform 
 ➔ GLM, GBM, RF, K-Means, PCA, Deep Learning
 Easy to use SDK / API
 ➔ Java, R, Scala, Python, JSON, Browser-based GUI ! Businesses can use ALL of their data (w or w/o Hadoop)
 ➔ Modeling without Sampling
 
 Big Data + Better Algorithms 
 ➔ Better Predictions H2O Open Source in-memory
 Prediction Engine for Big Data
  5. 5. About H20 (aka 0xdata) Pure Java, Apache v2 Open Source Join the www.h2o.ai/community!
  6. 6. H2O w or w/o Hadoop H2O H2O H2O HDFS HDFS HDFS YARN Hadoop MR R Java Scala JSON Python Standalone Over YARN On MRv1
  7. 7. H2O Architecture in-memory K-V store MapReduce compression Machine Learning Algorithms R Engine Nano fast Scoring Engine Prediction Engine memory manager e.g. Deep Learning
  8. 8. Wikipedia:
 Deep learning is a set of algorithms in machine learning that attempt to model high-level abstractions in data by using architectures composed of multiple non-linear transformations. ! ! ! ! ! Facebook DeepFace (LeCun): “Almost as good as humans at recognising faces” ! Google Brain (Andrew Ng, Jeff Dean & Geoffrey Hinton) ! FBI FACE: $1 billion face recognition project What is Deep Learning? Example: Input data
 (facial image) Prediction (person’s ID)
  9. 9. Deep Learning is trending 20132012 Google trends 2011
  10. 10. 1970s multi-layer feed-forward Neural Network (supervised learning with back-propagation) ! + distributed processing for big data (H2O in-memory MapReduce paradigm on distributed data) ! + multi-threaded speedup (H2O Fork/Join worker threads update the model asynchronously) ! + smart algorithms for accuracy (weight initialization, adaptive learning, momentum, dropout, regularization) ! = Top-notch prediction engine! Deep Learning in H2O
  11. 11. “fully connected” directed graph of neurons age income employment married not married Input layer Hidden layer 1 Hidden layer 2 Output layer 3x4 4x3 3x2#connections information flow input/output neuron hidden neuron 4 3 2#neurons 3 Example Neural Network
  12. 12. age income employment yj = tanh(sumi(xi*uij)+bj) uij xi yj per-class probabilities
 sum(pl) = 1 zk = tanh(sumj(yj*vjk)+ck) vjk zk pl pl = softmax(sumk(zk*wkl)+dl) wkl softmax(xk) = exp(xk) / sumk(exp(xk)) “neurons activate each other via weighted sums” Prediction: Forward Propagation married not married activation function: tanh alternative:
 x -> max(0,x) “rectifier” pl is a non-linear function of xi: can approximate ANY function with enough layers! bj, ck, dl: bias values
 (indep. of inputs)
  13. 13. age income employment xi standardize input xi: mean = 0, stddev = 1 ! horizontalize categorical variables, e.g. {full-time, part-time, none, self-employed} 
 ->
 {0,1,0} = part-time, {0,0,0} = self-employed Poor man’s initialization: random weights ! Better: Uniform distribution in
 +/- sqrt(6/(#units + #units_previous_layer)) Data preparation & Initialization Neural Networks are sensitive to numerical noise,
 operate best in the linear regime (not saturated) married not married
  14. 14. Mean Square Error = (0.2^2 + 0.2^2)/2 “penalize differences per-class” ! Cross-entropy = -log(0.8) “strongly penalize non-1-ness” Stochastic Gradient Descent SGD: improve weights and biases for EACH training row married not married For each training row, we make a prediction and compare with the actual label (supervised training): 1 0 0.8 0.2 predicted actual Objective: minimize prediction error (MSE or cross-entropy) w <— w - rate * ∂E/∂w 1
  15. 15. Backward Propagation 
 ! ∂E/∂wi = ∂E/∂y * ∂y/∂net * ∂net/∂wi = ∂(error(y))/∂y * ∂(activation(net))/∂net * xi Backprop: Compute ∂E/∂wi via chain rule going backwards wi net = sumi(wi*xi) + b xi E = error(y) y = activation(net) How to compute ∂E/∂wi for wi <— wi - rate * ∂E/∂wi ? Naive: For every i, evaluate E twice at (w1,…,wi±∆,…,wN)… Slow!
  16. 16. H2O Deep Learning Architecture K-V K-V HTTPD HTTPD nodes/JVMs: sync threads: async communication w w w w w w w w1 w3 w2 w4 w2+w4 w1+w3 w* = (w1+w2+w3+w4)/4 map:
 each node trains a copy of the weights and biases with (some* or all of) its local data with asynchronous F/J threads initial weights and biases w updated weights and biases w* H2O atomic in-memory
 K-V store reduce:
 average weights and biases from all nodes Keep iterating over the data (“epochs”), score from time to time Query & display the model via JSON, WWW 2 2 431 1 1 1 4 3 2 1 2 1 i *mini-batch: number of total rows per iteration, can be less than 1 epoch
  17. 17. “Secret” Sauce to Higher Accuracy Momentum training:
 keep changing weights and biases (even if there’s no error) 
 “find other local minima, and go faster along valleys” Adaptive learning rate - ADADELTA (Google):
 automatically set learning rate for each neuron based on its training history, combines annealing and momentum features Learning rate annealing:
 rate r = r0 / (1+ß*N), N = training samples “dig deeper into local minimum” Grid Search and Checkpointing:
 Run a grid search over multiple hyper-parameters, then continue training the best model L1/L2/Dropout/MaxSumWeights regularization:
 L1: penalizes non-zero weights, L2: penalizes large weights
 Dropout: randomly ignore certain inputs “train exp. many models at once” MaxSumWeights: Reduce all incoming weights if the sum > max value “regularization avoids overtraining and improves generalization error”
  18. 18. MNIST: digits classification Train: 60,000 rows 784 integer columns 10 classes Test: 10,000 rows 784 integer columns 10 classes MNIST: Digitized handwritten digits database (Yann LeCun) Data: 28x28=784 pixels with values in 0…255 (gray-scale) One of the most popular multi-class classification problems Without distortions or convolutions (which help), the best-ever published error rate on test set: 0.83% (Microsoft)
  19. 19. most frequent mistakes:
 confuse 4 with 6 and 9, and 7 with 2 test set error: 1.5% after 40 epochs 1.02% after 400 epochs
 0.95% after 4000 epochs H2O Deep Learning on MNIST: 0.95% test set error (so far) 1 node
  20. 20. Prostate Cancer Dataset
  21. 21. Live Demo: Cancer Prediction Interactive ROC curve with real- time updates
  22. 22. Live Demo: Cancer Prediction 0% training error with only 322 model parameters in seconds!
  23. 23. Live Demo: Grid Search Regression Doing a grid search to find good hyper-parameters to predict AGE from other 7 features Then continue training the best model 5 hidden 50 tanh layers, rho=0.99, epsilon = 1e-10 MSE < 1 for test set ages in 44…79 Regression: 1 linear output neuron
  24. 24. Live Demo: ebay Text Classification Users enter a description when selling an item Task: Predict the type of item Data prep: Binary word vector 0,0,1,0,0,0,0,0,1,0,0,0,1,…,0 H2O parses SVMLight sparse format: label 3:1 9:1 13:1 … ! “Small” sample dataset on jewelry and watches: Train: 578,361 rows 8,647 cols 467 classes Test: 64,263 rows 8,647 cols 143 classes ! H2O compressed columnar in-memory store: Only needs 60MB to store 5 billion entries (never inflated)
  25. 25. Live Demo: ebay Text Classification Work in progress, shown results are for illustration only! Default parameters, no tuning, 4 nodes (16-cores each) Train: 578,361 rows 8,647 cols 467 classes Test: 64,263 rows 8,647 cols 143 classes
  26. 26. Tips for H2O Deep Learning ! General: More layers: more complex functions (non-linearity) More neurons per layer: detect finer structure in data More regularization: less overfitting (better validation error) ! Do a grid search to get a feel for convergence, then continue training. Try Tanh first. For Rectifier, try max_w2 = 50 and/or L1=1e-5. Try TanhDropout or RectifierDropout with test/validation set after finding good parameters for convergence on training set. Distributed: Smaller mini-batch: more comm., slower, but higher accuracy. With ADADELTA: Try epsilon = 1e-4,1e-6,1e-8,1e-10, rho = 0.9,0.95,0.99 Without ADADELTA: Try rate = 1e-4…1e-2, rate_annealing = 1e-5…1e-8 Try momentum_start = 0.5, momentum_stable = 0.99,
 momentum_ramp = 1/rate_annealing Try balance_classes = true for imbalanced classes. Try force_load_balance for small datasets.
  27. 27. Summary H2O is a distributed in-memory math platform that allows fast prototyping in Java, R, Scala and Python. ! H2o enables the development of enterprise-quality blazing fast machine learning applications. ! H2O Deep Learning is distributed, easy to use, and early results compete with the world’s best. ! Deep Learning makes better predictions! ! Try it yourself and join our next meetup!
 git clone https://github.com/0xdata/h2o

×