Software Engineering for Machine Learning

A draft book by Yung-Hsiang Lu and and George K. Thiruvathukal.

• Preface
  • The Need of Integration
  • Principles in Writing this Book
  • Audience and Expected Background
  • Open-Source Book
  • About the Authors
• Introduction to Python
  • Characteristics of Python
  • Executing Programs
  • Arithmetic Expressions
  • Built-in Arithmetic Functions
  • Assignments and Variables
  • Creating Functions
  • Modules
  • Files
  • print() and printing
  • input() and raw_input()
  • while loops
  • Lists
  • for loops
  • continue Statement
  • break and else in loops
  • List Comprehensions
  • Dictionary Comprehensions
  • None
  • More on Assignment
  • Dictionaries
  • Strings
• Version Control
  • Software Versions
  • github
  • Clone a Repository
  • Commit and Push
  • Add and Remove Files
  • Collaboration using github
  • Branches
  • Pull Requests
  • Issues
  • Release this Book
  • Set Up SSH Key
• Code Review
  • Modern Code Review
  • How To Start A Code Review
  • What to Review
• Testing
  • Structure of Tests
  • Use Assert Correctly
  • Exception
  • Pytest
  • Test Integer Partition
  • Test Coverage
  • Continous Integration
  • Limitations of Testing
  • Halting Problems: Unsolvable Problems
• Web Services
• Continuous Integration
• Graphics
• Reproducibility
  • Reproducibility Packages
  • Containers
  • Zenodo and GitHub Release Archival
• Machine Learning
  • Decision Problems
  • Decisions and Feedback
  • Knowledge in Data
  • Define Learning
  • Limitations of Machine Learning
  • Black Swan Events
• K-Means Clustering
  • Clustering Data
  • Clustering into Groups
  • Define K-Mean Clustering Problem
  • Number of Solutions for Clustering
  • K-Mean Algorithm
  • Generate Test Cases
  • Visualize Data
  • K-Mean Implementation
  • Understand Non-Deterministic Behavior of K-Mean
  • Determine k’s Value
  • Clean Data before Clustering
  • Summary
• Hierarchical Clustering
  • Limitations of the k-mean Algorithm
  • Examples of Hierarchical clustering
  • Hierarchical Clustering Algorithm
  • Define Distance of Two Clusters
  • Implementation
• Principal Component Analysis
  • Examples
• Practical Considerations of Clustering
• Floating Point and Finite Precision
• Supervised Machine Learning
  • Distinguish Supervised and Unsupervised Learning
  • Minimization Problem
  • Classification and Regression Problems
  • Quantify Errors in Classification
  • Regression Problems
  • Overfitting
• Gradient Descent
  • Linear Approximation with Least Square Error
  • Gradient
  • Gradient Descent
  • Gradient Descent for Neural Networks
• Support Vector Machines
• Neural Networks
• Building the Book using Sphinx, GitHub Pages, and Travis
• Sphinx Demonstration

Repository

github.com/PurdueCAM2Project/SE4ML

Continuous Integration

travis-ci.com/PurdueCAM2Project/SE4ML

Download Release Builds

Available as printable PDF and eBook

Cite

Yung-Hsiang Lu and George K. Thiruvathukal, Software Engineering and Machine Learning, PurdueCAM2Project/SE4ML, Zenodo, http://doi.org/10.5281/zenodo.2532051.

Topics

Index