Hyperparameter Tuning with Python

by Louis Owen

Data Science

Book Details

Book Title

Hyperparameter Tuning with Python

Author

Louis Owen

Publisher

Packt

Publication Date

2022

ISBN

9781803235875

Number of Pages

306

Language

English

Format

PDF

File Size

4MB

Subject

System Design

Table of Contents

  • Cover
  • Title Page
  • Copyright and Credits
  • Contributors
  • Table of Contents
  • Preface
  • Section 1: The Methods
  • Chapter 1: Evaluating Machine Learning Models
  • Technical requirements
  • Understanding the concept of overfitting
  • Creating training, validation, and test sets
  • Exploring random and stratified splits
  • Discovering repeated k-fold cross-validation
  • Discovering Leave-One-Out cross-validation
  • Discovering LPO cross-validation
  • Discovering time-series cross-validation
  • Summary
  • Further reading
  • Chapter 2: Introducing Hyperparameter Tuning
  • What is hyperparameter tuning?
  • Demystifying hyperparameters versus parameters
  • Understanding hyperparameter space and distributions
  • Summary
  • Chapter 3: Exploring Exhaustive Search
  • Understanding manual search
  • Understanding grid search
  • Understanding random search
  • Summary
  • Chapter 4: Exploring Bayesian Optimization
  • Introducing BO
  • Understanding BO GP
  • Understanding SMAC
  • Understanding TPE
  • Understanding Metis
  • Summary
  • Chapter 5: Exploring Heuristic Search
  • Understanding simulated annealing
  • Understanding genetic algorithms
  • Understanding particle swarm optimization
  • Understanding Population-Based Training
  • Summary
  • Chapter 6: Exploring Multi-Fidelity Optimization
  • Introducing MFO
  • Understanding coarse-to-fine search
  • Understanding successive halving
  • Understanding hyper band
  • Understanding BOHB
  • Summary
  • Section 2: The Implementation
  • Chapter 7: Hyperparameter Tuning via Scikit
  • Technical requirements
  • Introducing Scikit
  • Implementing Grid Search
  • Implementing Random Search
  • Implementing Coarse-to-Fine Search
  • Implementing Successive Halving
  • Implementing Hyper Band
  • Implementing Bayesian Optimization Gaussian Process
  • Implementing Bayesian Optimization Random Forest
  • Implementing Bayesian Optimization Gradient Boosted Trees
  • Summary
  • Chapter 8: Hyperparameter Tuning via Hyperopt
  • Technical requirements
  • Introducing Hyperopt
  • Implementing Random Search
  • Implementing Tree-structured Parzen Estimators
  • Implementing Adaptive TPE
  • Implementing simulated annealing
  • Summary
  • Chapter 9: Hyperparameter Tuning via Optuna
  • Technical requirements
  • Introducing Optuna
  • Implementing TPE
  • Implementing Random Search
  • Implementing Grid Search
  • Implementing Simulated Annealing
  • Implementing Successive Halving
  • Implementing Hyperband
  • Summary
  • Chapter 10: Advanced Hyperparameter Tuning with DEAP and Microsoft NNI
  • Technical requirements
  • Introducing DEAP
  • Implementing the Genetic Algorithm
  • Implementing Particle Swarm Optimization
  • Introducing Microsoft NNI
  • Implementing Grid Search
  • Implementing Random Search
  • Implementing Tree-structured Parzen Estimators
  • Implementing Sequential Model Algorithm Configuration
  • Implementing Bayesian Optimization Gaussian Process
  • Implementing Metis
  • Implementing Simulated Annealing
  • Implementing Hyper Band
  • Implementing Bayesian Optimization Hyper Band
  • Implementing Population-Based Training
  • Summary
  • Section 3: Putting Things into Practice
  • Chapter 11: Understanding the Hyperparameters of Popular Algorithms
  • Exploring Random Forest hyperparameters
  • Exploring XGBoost hyperparameters
  • Exploring LightGBM hyperparameters
  • Exploring CatBoost hyperparameters
  • Exploring SVM hyperparameters
  • Exploring artificial neural network hyperparameters
  • Summary
  • Chapter 12: Introducing Hyperparameter Tuning Decision Map
  • Getting familiar with HTDM
  • Case study 1 – using HTDM with a CatBoost classifier
  • Case study 2 – using HTDM with a conditional hyperparameter space
  • Case study 3 – using HTDM with prior knowledge of the hyperparameter values
  • Summary
  • Chapter 13: Tracking Hyperparameter Tuning Experiments
  • Technical requirements
  • Revisiting the usual practices
  • Exploring Neptune
  • Exploring scikit-optimize
  • Exploring Optuna
  • Exploring Microsoft NNI
  • Exploring MLflow
  • Summary
  • Chapter 14: Conclusions and Next Steps
  • Revisiting hyperparameter tuning methods and packages
  • Revisiting HTDM
  • What’s next?
  • Summary
  • Index
  • Other Books You May Enjoy