(NOTES) Structuring Machine Learning Projects

3. Structuring Machine Learning Projects:

Slides

3.1 Week 1 - Introduction to ML Strategy / Setting Up your Goal / Comparing to Human-level Performance

3.1.1 Why ML Strategy?

3.1.2 Orthogonalization

3.1.3 Single Number Evaluation Metric

3.1.4 Satisficing and Optimizing Metric

3.1.5 Train/Dev/Test Distributions

3.1.6 Size of the Dev and Test Sets

3.1.7 When to Change Dev/Test Sets and Metrics?

3.1.8 Why Human-level Performance?

3.1.9 Avoidable Bias

3.1.10 Understanding Human-level Performance

3.1.11 Surpassing Human-level Performance

3.1.12 Improving your Model Performance

3.2 Week 2 - Error Analysis / Mismatched Training and Dev/Test Set / Learning from Multiple Tasks / End-to-end Deep Learning

3.2.1 Carrying Out Error Analysis

3.2.2 Cleaning Up Incorrectly Labeled Data

3.2.3 Build your First System Quickly, then Iterate

3.2.4 Training and Testing on Different Distributions

3.2.5 Bias and Variance with Mismatched Data Distributions

3.2.6 Addressing Data Mismatch

3.2.7 Transfer Learning

3.2.8 Multi-task Learning

3.2.9 What is End-to-end Deep Learning?

3.2.10 Whether to use End-to-end Deep Learning