Learn how to set up various prediction endpoints using vetiver for a model predicting Bigfoot sightings in this 29-minute screencast. Explore the process of feature engineering, natural language processing, and logistic regression as you analyze #TidyTuesday Bigfoot sighting data. Discover techniques for tuning, evaluating model performance through confusion matrices, and assessing variable importance. Gain practical insights into deploying the model and making predictions, with accompanying code available on the presenter's blog for further study and implementation.
Overview
Syllabus
Introduction
Class C secondhand reports
Class A secondhand reports mutate
Exploratory work
Data analysis
Feature engineering
Natural language
Logistic regression
Tuning grid
Show best
Last fit
Confusion Matrix
Variable Importance
Deployment
Predict
Summary
Taught by
Julia Silge
Related Courses
-
Building and Evaluating Text Classifiers in Python
-
AI skills for Engineers: Supervised Machine Learning
-
Understand data science for machine learning
-
Logistic Regression Implementation in R for Automotive Crash Testing
-
Feature Engineering & Interpretability for XGBoost with Board Game Ratings
-
Tune and Evaluate a Multiclass Lasso Model for NBER Working Papers
