Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

Learning to Predict User Preferences

Go to class Write review

Overview

Take your recommendation engine to the next level with supervised machine learning! You'll learn how to prepare training data from user listening logs, train a classification model (like Logistic Regression) to predict track affinity, and use these predictions to offer even more personalized music suggestions.

Syllabus

Unit 1: Preparing Training Data

Preparing Training Data from Session Logs Example
Correct the Data Prep Logic to Reflect Real User Preferences
Preparing Training Data from Session Logs Example
Build the Core Function from Guided Skeleton

Unit 2: Training Track Affinity Model

Training a Logistic Regression Model for Predicting User-Track Affinity
Ensure ROC AUC is Calculated on the Correct Dataset
Write Tests to Validate Affinity Model Training Logic
Implement the Track Affinity Model Trainer

Unit 3: Scoring Tracks with Confidence

Observe and Understand: Predicting Track Affinity Scores
Scoring Tracks with Predictive Confidence Example
Scoring Tracks with Predictive Confidence Example
Test Predictive Scoring Logic with and without a Model

Unit 4: Building Predictive Recommendation API

Explore the Predictive Recommendation API Endpoint
Implementing the Predictive Recommendation API Endpoint
Test the Predictive Recommendation API

Reviews

Start your review of Learning to Predict User Preferences