Learn to build, tune, and evaluate a multiclass lasso model for predicting economic research paper categories using #TidyTuesday data. Explore the NBER working papers dataset, analyze weighted log odds, and implement feature engineering techniques. Master preprocessing steps, model specification, and hyperparameter tuning. Gain hands-on experience in making predictions and interpreting results. Follow along with Julia Silge's comprehensive tutorial, which includes code examples and in-depth explanations of each step in the process.
Overview
Syllabus
Introduction
Data set overview
Weighted log odds
Highest log odds words
Multiclass classification
Feature engineering
Preprocessing
Model specification
Putting it together
Tuning
Results
Lastfit
Making predictions
Taught by
Julia Silge
