Recommendations for Building Machine Learning Software
Download as PPTX, PDF18 likes2,960 views
This document provides recommendations for building machine learning software from the perspective of Netflix's experience. The first recommendation is to be flexible about where and when computation happens by distributing components across offline, nearline, and online systems. The second is to think about distribution starting from the outermost levels of the problem by parallelizing across subsets of data, hyperparameters, and machines. The third recommendation is to design application software for experimentation by sharing components between experiment and production code. The fourth recommendation is to make algorithms and models extensible and modular by providing reusable building blocks. The fifth recommendation is to describe input and output transformations with models. The sixth recommendation is to not rely solely on metrics for testing and instead implement unit testing of code.
![28
Example: Simple ranking system
High-level API: List<Video> rank(User u, List<Video> videos)
Example model description file:
{
“type”: “ScoringRanker”,
“scorer”: {
“type”: “FeatureScorer”,
“features”: [
{“type”: “Popularity”, “days”: 10},
{“type”: “PredictedRating”}
],
“function”: {
“type”: “Linear”,
“bias”: -0.5,
“weights”: {
“popularity”: 0.2,
“predictedRating”: 1.2,
“predictedRating*popularity”:
3.5
}
}
}
Ranker
Scorer
Features
Linear function
Feature transformations](https://image.slidesharecdn.com/recommendationsformlsoftwarepublic-151114003739-lva1-app6891/85/Recommendations-for-Building-Machine-Learning-Software-28-320.jpg)
Recommendations for Building Machine Learning Software
- 1. 11 Recommendations for Building Machine Learning Software Justin Basilico Page Algorithms Engineering November 13, 2015 @JustinBasilico SF 2015
- 3. 3 Change of focus 2006 2015
- 4. 4 Netflix Scale > 69M members > 50 countries > 1000 device types > 3B hours/month 36.4% of peak US downstream traffic
- 5. 5 Goal Help members find content to watch and enjoy to maximize member satisfaction and retention
- 6. 6 Everything is a Recommendation Rows Ranking Over 80% of what people watch comes from our recommendations Recommendations are driven by Machine Learning
- 8. 8 Models & Algorithms Regression (linear, logistic, elastic net) SVD and other Matrix Factorizations Factorization Machines Restricted Boltzmann Machines Deep Neural Networks Markov Models and Graph Algorithms Clustering Latent Dirichlet Allocation Gradient Boosted Decision Trees/Random Forests Gaussian Processes …
- 9. 9 Design Considerations Recommendations • Personal • Accurate • Diverse • Novel • Fresh Software • Scalable • Responsive • Resilient • Efficient • Flexible
- 12. 12 Be flexible about where and when computation happens Recommendation 1
- 13. 13 System Architecture Offline: Process data Batch learning Nearline: Process events Model evaluation Online learning Asynchronous Online: Process requests Real-time Netflix.Hermes Netflix.Manhattan Nearline Computation Models Online Data Service Offline Data Model training Online Computation Event Distribution User Event Queue Algorithm Service UI Client Member Query results Recommendations NEARLINE Machine Learning Algorithm Machine Learning Algorithm Offline Computation Machine Learning Algorithm Play, Rate, Browse... OFFLINE ONLINE More details on Netflix Techblog
- 14. 14 Where to place components? Example: Matrix Factorization Offline: Collect sample of play data Run batch learning algorithm like SGD to produce factorization Publish video factors Nearline: Solve user factors Compute user-video dot products Store scores in cache Online: Presentation-context filtering Serve recommendations Netflix.Hermes Netflix.Manhattan Nearline Computation Models Online Data Service Offline Data Model training Online Computation Event Distribution User Event Queue Algorithm Service UI Client Member Query results Recommendations NEARLINE Machine Learning Algorithm Machine Learning Algorithm Offline Computation Machine Learning Algorithm Play, Rate, Browse... OFFLINE ONLINE V sij=uivj Aui=b sij X≈UVt X sij>t
- 15. 15 Think about distribution starting from the outermost levels Recommendation 2
- 16. 16 Three levels of Learning Distribution/Parallelization 1. For each subset of the population (e.g. region) Want independently trained and tuned models 2. For each combination of (hyper)parameters Simple: Grid search Better: Bayesian optimization using Gaussian Processes 3. For each subset of the training data Distribute over machines (e.g. ADMM) Multi-core parallelism (e.g. HogWild) Or… use GPUs
- 17. 17 Example: Training Neural Networks Level 1: Machines in different AWS regions Level 2: Machines in same AWS region Spearmint or MOE for parameter optimization Mesos, etc. for coordination Level 3: Highly optimized, parallel CUDA code on GPUs
- 18. 18 Design application software for experimentation Recommendation 3
- 19. 19 Example development process Idea Data Offline Modeling (R, Python, MATLAB, …) Iterate Implement in production system (Java, C++, …) Data discrepancies Missing post- processing logic Performance issues Actual output Experimentation environment Production environment (A/B test) Code discrepancies Final model
- 20. 20 Shared Engine Avoid dual implementations Experiment code Production code ProductionExperiment • Models • Features • Algorithms • …
- 21. 21 Solution: Share and lean towards production Developing machine learning is iterative Need a short pipeline to rapidly try ideas Want to see output of complete system So make the application easy to experiment with Share components between online, nearline, and offline Use the real code whenever possible Have well-defined interfaces and formats to allow you to go off-the-beaten-path
- 22. 22 Make algorithms extensible and modular Recommendation 4
- 23. 23 Make algorithms and models extensible and modular Algorithms often need to be tailored for a specific application Treating an algorithm as a black box is limiting Better to make algorithms extensible and modular to allow for customization Separate models and algorithms Many algorithms can learn the same model (i.e. linear binary classifier) Many algorithms can be trained on the same types of data Support composing algorithms Data Parameters Data Model Parameters Model Algorithm Vs.
- 24. 24 Provide building blocks Don’t start from scratch Linear algebra: Vectors, Matrices, … Statistics: Distributions, tests, … Models, features, metrics, ensembles, … Loss, distance, kernel, … functions Optimization, inference, … Layers, activation functions, … Initializers, stopping criteria, … … Domain-specific components Build abstractions on familiar concepts Make the software put them together
- 25. 25 Example: Tailoring Random Forests Using Cognitive Foundry: http://github.com/algorithmfoundry/Foundry Use a custom tree split Customize to run it for an hour Report a custom metric each iteration Inspect the ensemble
- 26. 26 Describe your input and output transformations with your model Recommendation 5
- 27. 27 Application Putting learning in an application Feature Encoding Output Decoding ? Machine Learned Model Rd ⟶ Rk Application or model code?
- 28. 28 Example: Simple ranking system High-level API: List<Video> rank(User u, List<Video> videos) Example model description file: { “type”: “ScoringRanker”, “scorer”: { “type”: “FeatureScorer”, “features”: [ {“type”: “Popularity”, “days”: 10}, {“type”: “PredictedRating”} ], “function”: { “type”: “Linear”, “bias”: -0.5, “weights”: { “popularity”: 0.2, “predictedRating”: 1.2, “predictedRating*popularity”: 3.5 } } } Ranker Scorer Features Linear function Feature transformations
- 29. 29 Don’t just rely on metrics for testing Recommendation 6
- 30. 30 Machine Learning and Testing Temptation: Use validation metrics to test software When things work and metrics go up this seems great When metrics don’t improve was it the code data metric idea …?
- 31. 31 Reality of Testing Machine learning code involves intricate math and logic Rounding issues, corner cases, … Is that a + or -? (The math or paper could be wrong.) Solution: Unit test Testing of metric code is especially important Test the whole system: Just unit testing is not enough At a minimum, compare output for unexpected changes across versions
- 32. 3232 Conclusions
- 33. 33 Two ways to solve computational problems Know solution Write code Compile code Test code Deploy code Know relevant data Develop algorithmic approach Train model on data using algorithm Validate model with metrics Deploy model Software Development Machine Learning (steps may involve Software Development)
- 34. 34 Take-aways for building machine learning software Building machine learning is an iterative process Make experimentation easy Take a holistic view of application where you are placing learning Design your algorithms to be modular Look for the easy places to parallelize first Testing can be hard but is worthwhile
- 35. 35 Thank You Justin Basilico [email protected] @JustinBasilico We’re hiring