Downloaded 34 times
Uploaded bySri Ambati
Arno candel scalabledatascienceanddeeplearningwithh2o_reworkboston2015
The document presents a workshop by Arno Candel from h2o.ai, focusing on scalable data science and deep learning applications. It covers h2o.ai's machine intelligence offerings, architecture, live demos on various datasets, and their cutting-edge algorithms for efficient data processing. The session highlights real-world applications, competitive advantages, and the open-source nature of their tools for machine learning practitioners.
More Related Content
Similar to Arno candel scalabledatascienceanddeeplearningwithh2o_reworkboston2015
Arno candel scalabledatascienceanddeeplearningwithh2o_reworkboston2015
- 1. H2O.ai Machine Intelligence Scalable DataScience and Deep Learning with H2O 1 RE:WORK Deep Learning Summit Workshop Boston, May 26 2015 Arno Candel, PhD Chief Architect, H2O.ai
- 2. H2O.ai Machine Intelligence Who AmI? Arno Candel Chief Architect, Physicist & Hacker at H2O.ai PhD Physics, ETH Zurich 2005 10+ yrs Supercomputing (HPC) 6 yrs at SLAC (Stanford Lin. Accel.) 3.5 yrs Machine Learning 1.5 yrs at H2O.ai Fortune Magazine Big Data All Star 2014 Follow me @ArnoCandel 2
- 3. H2O.ai Machine Intelligence Outline • Introduction •H2O Deep Learning Architecture • Live Demos: Flow GUI - Airline Dataset R - MNIST World Record + Anomaly Detection Flow GUI - Higgs Boson Classification Sparkling Water - Chicago Crime Prediction iPython - CitiBike Demand Prediction Scoring Engine - Million Songs Classification • Outlook 3
- 4. H2O.ai Machine Intelligence In-Memory ML Distributed OpenSource APIs 4 Memory-Efficient Data Structures Cutting-Edge Algorithms Use all your Data (No Sampling) Accuracy with Speed and Scale Ownership of Methods - Apache V2 Easy to Deploy: Bare, Hadoop, Spark, etc. Java, Scala, R, Python, JavaScript, JSON NanoFast Scoring Engine (POJO) H2O - Product Overview
- 5. H2O.ai Machine Intelligence 25,000 commits/ 3yrs H2O World Conference 2014 Team Work @ H2O.ai 5 Join H2O World 2015!
- 6. H2O.ai Machine Intelligence 103 6342789 463 2,887 13,237 Companies Users Mar 2014 July 2014 Mar 2015 Active Users 150+ 6 Strong Community & Growth 5/25/15 @kdnuggets t.co/4xSgleSIdY
- 7. H2O.ai Machine Intelligence 7 HDFS S3 SQL NoSQL Classification Regression Feature Engineering Distributed In-Memory MapReduce/Fork Join Columnar Compression GLM, Deep Learning K-Means, PCA, NB, Cox Random Forest / GBM Ensembles Fast Modeling Engine Streaming Nano Fast Java Scoring Engines (POJO code generation) Matrix Factorization Clustering Munging Unsupervised Supervised Accuracy with Speed and Scale Most code is written in-house from scratch
- 8. H2O.ai Machine Intelligence 8 Ad Optimization(200% CPA Lift with H2O) P2B Model Factory (60k models, 15x faster with H2O than before) Fraud Detection (11% higher accuracy with H2O Deep Learning - saves millions) …and many large insurance and financial services companies! Real-time marketing (H2O is 10x faster than anything else) Actual Customer Use Cases
- 9.
- 10. H2O.ai Machine Intelligence 10 Airline Data:Predict Delayed Departure Predict dep. delay Y/N 116M rows 31 colums 12 GB CSV 4 GB compressed 20 years of domestic airline flight data
- 11. H2O.ai Machine Intelligence 11 Results inSeconds on Big Data Logistic Regression: ~20s elastic net, alpha=0.5, lambda=1.379e-4 (auto) Deep Learning: ~70s 4 hidden ReLU layers of 20 neurons, 2 epochs 8-node EC2 cluster: 64 virtual cores, 1GbE Year, Month, Sched. Dep. Time have non-linear impact Chicago, Atlanta, Dallas: often delayed All cores maxed out +9% AUC +--+++
- 12. H2O.ai Machine Intelligence Multi-layer feed-forwardNeural Network Trained with back-propagation (SGD, ADADELTA) + distributed processing for big data (fine-grain in-memory MapReduce on distributed data) + multi-threaded speedup (async fork/join worker threads operate at FORTRAN speeds) + smart algorithms for fast & accurate results (automatic standardization, one-hot encoding of categoricals, missing value imputation, weight & bias initialization, adaptive learning rate, momentum, dropout/l1/L2 regularization, grid search, N-fold cross-validation, checkpointing, load balancing, auto-tuning, model averaging, etc.) = powerful tool for (un)supervised machine learning on real-world data 12 H2O Deep Learning all 320 cores maxed out
- 13. H2O.ai Machine Intelligence threads: async 13 H2ODeep Learning Architecture K-V K-V HTTPD HTTPD nodes/JVMs: sync 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 model: weights and biases w updated model: w* H2O in-memory non- blocking hash map: K-V store reduce: model averaging: average weights and biases from all nodes, speedup is at least #nodes/ log(#rows) http://arxiv.org/abs/1209.4129 Keep iterating over the data (“epochs”), score at user-given times Query & display the model via JSON, WWW 2 2 4 31 1 1 1 4 3 2 1 2 1 i *auto-tuned (default) or user- specified number of rows per MapReduce iteration Main Loop:
- 14. H2O.ai Machine Intelligence 14 H2O DeepLearning beats MNIST MNIST: Handwritten digits: 28^2=784 gray-scale pixel values full run: 10 hours on 10-node cluster 2 hours on desktop gets to 0.9% test set error Just supervised training on original 60k/10k dataset: No data augmentation No distortions No convolutions No pre-training No ensemble 0.83% test set error: current world record 1-liner: call h2o.deeplearning() in R
- 15. H2O.ai Machine Intelligence 15 Unsupervised AnomalyDetection The good The bad The ugly Try it yourself! Auto-Encoder learns “Compressed Identity”
- 16. H2O.ai Machine Intelligence 16 Images courtesyCERN / LHC Higgs vs Background Large Hadron Collider: Largest experiment of mankind! $13+ billion, 16.8 miles long, 120 MegaWatts, -456F, 1PB/day, etc. Higgs boson discovery (July ’12) led to 2013 Nobel prize! Higgs Boson - Classification Problem
- 17. H2O.ai Machine Intelligence 17 UCI HiggsDataset: 11M rows, 29 cols C2-C22: 21 low-level features (detector data) 7 high-level features (physics formulae) Assume we don’t know Physics…
- 18. H2O.ai Machine Intelligence 18 ? ?? Former CERN baseline for AUC: 0.733 and 0.816 H2O Algorithm low-level H2O AUC all features H2O AUC Generalized Linear Model 0.596 0.684 Random Forest 0.764 0.840 Gradient Boosted Trees 0.753 0.839 Neural Net 1 hidden layer 0.760 0.830 H2O Deep Learning ? add derived features Deep Learning for Higgs Detection? Q: Can Deep Learning learn Physics for us?
- 19. H2O.ai Machine Intelligence 19 EC2 DemoCluster: 8 nodes, 64 cores H2O Deep Learning: Expect good cluster utilization :)
- 20. H2O.ai Machine Intelligence 20 Deep DLmodel on low-level features only valid 500k rows test 500k rows train 10M rows H2O Deep Learning Higgs Demo H2O: same results as Nature paper Deep Learning just learned Particle Physics! 8 EC2 nodes: AUC = 0.86 after 100 mins AUC = 0.87+ overnight
- 21.
- 22. H2O.ai Machine Intelligence 22 Weather +Census + Crime Data
- 23. H2O.ai Machine Intelligence 23 Spark+ H2O = Sparkling Water
- 24. H2O.ai Machine Intelligence 24 Sparkling WaterDemo Instructions at h2o.ai/download
- 25. H2O.ai Machine Intelligence 25 Parse &Munge with H2O, Convert to RDD H2O Parser: Robust & Fast Simple Column Selection
- 26. H2O.ai Machine Intelligence 26 Parse &Munge with H2O, Convert to RDD Munging: Date Manipulations Conversion to DataFrame
- 27. H2O.ai Machine Intelligence 27 Join RDDswith SQL, Convert to H2O Spark SQL Query Execution Convert back to H2OFrame Split into Train 80% / Test 20%
- 28. H2O.ai Machine Intelligence 28 Build H2ODeep Learning Model Train a H2O Deep Learning Model on Data obtained by Spark SQL Query Predict whether Arrest will be made with AUC of 0.90+
- 29. H2O.ai Machine Intelligence 29 Visualize Resultswith Flow Using Flow to interactively plot Arrest Rate (blue) vs Relative Occurrence (red) per crime type.
- 30. H2O.ai Machine Intelligence 30 Predict RentalBike Demand in NYC Cliff et al.
- 31.
- 32. H2O.ai Machine Intelligence iPython Notebook Demo 32 ModelBuilding And Scoring 91% AUC baseline
- 33.
- 34. H2O.ai Machine Intelligence 34 Improved Modelswith Weather Data 93% AUC after joining bike and weather data
- 35. H2O.ai Machine Intelligence 35 Example: FirstGBM tree Fast and easy path to Production (batch or real-time)! POJO Scoring Engine Standalone Java scoring code is auto-generated! Note: no heap allocation, pure decision-making
- 36. H2O.ai Machine Intelligence More Infoin H2O Booklets https://leanpub.com/u/h2oai http://learn.h2o.ai 36
- 37. H2O.ai Machine Intelligence 37 Kaggle -H2O Starter Scripts
- 38. H2O.ai Machine Intelligence 38 Kaggle -Active H2O Community 18k+ views in 2 weeks
- 39. H2O.ai Machine Intelligence 39 Hyper-Parameter Tuning 93numerical features 9 output classes 62k training set rows 144k test set rows Ensemble of H2O DL + GBM => Top 10% R script by DataGeek
- 40. H2O.ai Machine Intelligence 40 Mastering Kagglewith H2O DL + GBM GBM GBM + GLM DRF + GLM Stay tuned: Kaggle Master @Mark_A_Landry recently joined H2O as Competitive Data Scientist! www.meetup.com/Silicon- Valley-Big-Data-Science/ events/222303884/
- 41.
- 42. H2O.ai Machine Intelligence Outlook -Algorithm Roadmap • Ensembles (Erin LeDell et al.) • Automated Hyper-Parameter Tuning • Convolutional Layers for Deep Learning • Natural Language Processing: tf-idf, Word2Vec, … • Generalized Low Rank Models • PCA, SVD, K-Means, Matrix Factorization • Recommender Systems And many more! 42 Public JIRAs - Join H2O!
- 43. H2O.ai Machine Intelligence Key Take-Aways H2Ois an open source predictive analytics platform for data scientists and business analysts who need scalable, fast and accurate machine learning. H2O Deep Learning is ready to take your advanced analytics to the next level. Try it on your data! 43 https://github.com/h2oai H2O Google Group http://h2o.ai @h2oai Thank You!