Published Sep 11, 2015

The Model Complexity Myth

Jake VanderPlas challenges the "model complexity myth" by demystifying when fitting models with more parameters than data points is appropriate, while delving into regularization techniques, his contributions to open source projects, and the interplay of data science with astronomy through innovative tools like AstroML.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

The Computational Complexity of Machine Learning
Answers 383 questions
Complexity and Cryptography
Answers 383 questions
The Complexity of Learning Neural Networks
Answers 383 questions
Modeling Fake News
Answers 383 questions
Introduction
Answers 383 questions
Mathematical Models of Ecological Systems
Answers 383 questions
Machine Learning Done Wrong
Answers 383 questions
Data Myths with Karl Mamer
Answers 383 questions
[MINI] Exponential Time Algorithms
Answers 383 questions
Opinion Dynamics Models
Answers 383 questions
Oceanography and Data Science
Answers 383 questions
Game Theory
Answers 383 questions
Models of Mental Simulation
Answers 383 questions
The Experimental Design of Paranormal Claims
Answers 383 questions
Visualization and Interpretability
Answers 383 questions

Dexa/Kyle Polich

The Model Complexity Myth

Topics covered

Popular Clips

Understanding Regularization

Exoplanet Detection Methods

AstroML and Machine Learning

Matrix Inversion Challenges

Outlier Detection Techniques

Forward Modeling Insights

Compressed Sensing Explained

Model Complexity Explained

Stellar Brightness Modeling

Data Science Foundations

Episode Highlights

Jake VanderPlas's Contributions

Model Complexity

Data Science Techniques

Related Episodes

The Computational Complexity of Machine Learning

Complexity and Cryptography

The Complexity of Learning Neural Networks

Modeling Fake News

Introduction

Mathematical Models of Ecological Systems

Machine Learning Done Wrong

Data Myths with Karl Mamer

[MINI] Exponential Time Algorithms

Opinion Dynamics Models

Oceanography and Data Science

Game Theory

Models of Mental Simulation

The Experimental Design of Paranormal Claims

Visualization and Interpretability

The Model Complexity Myth

Topics covered

Popular Clips

Episode Highlights

Jake VanderPlas's ContributionsJake VanderPlas, an astronomer and data scientist, discusses his Python Data Science Handbook and its role in teaching data science. He also highlights his contributions to open source projects, emphasizing their significance in the data science community.

Jake VanderPlas's Contributions

Model Complexity

Data Science TechniquesJake VanderPlas explores the intricacies of regularization in data science, highlighting its role in managing model complexity and parameter estimation. He also discusses AstroML's impact on astronomy and the potential of compressed sensing in data reconstruction.

Data Science Techniques

Related Episodes