CSE 158/258 (MGTA461/DSC256): Web Mining and Recommender Systems
Autumn 2023, Tuesday/Thursday 11:00-12:20, WLH (Warren Lecture Hall) 2001 and Twitch

For those unable to access twitch, or attend the lecture time, all recordings will be posted below and to the UCSD Podcast page
Intro:
CSE 158 and 258 are undergraduate and graduate courses devoted to current methods for recommender systems, data mining, and predictive analytics. No previous background in machine learning is required, but all participants should be comfortable with programming (all example code will be in Python), and with basic optimization and linear algebra.

Lectures:
The course meets twice a week on Tuesday/Thursday mornings, starting September 28. The class meets in WLH (Warren Lecture Hall) 2001, though meetings will also be livestreamed on twitch. Recordings will also be made available on this page after each class.

Textbook:
The textbook for this course is Personalized Machine Learning . A (draft) pdf of the textbook is available for download .

Additional references are provided from Pattern Recognition and Machine Learning (Bishop) , and from Charles Elkan's 2013 course notes . Links are also provided to our Coursera Specialization , which covers similar (though more introductory) material.

Office hours:
Office hours (and instructions to access) for each class are posted to Piazza

Assessment:
Grading:
Each Homework is worth 8%. Your lowest (of four) homework grades is dropped (or one homework can be skipped).
The (take-home) Midterm is worth 25%.
Assignment 1 is worth 22%.
Assignment 2 is worth 25%.
Peer grading for Assignment 2 is worth 4%.
Assignment 2 is a group assignment . All other assessment must be completed individually.
All assessments are due before the Tuesday lecture on the due date. Late submissions are not accepted.
Wk 0/1 Supervised Learning: Regression

Least-squares regression
Overfitting and regularization
Feature Engineering
Training, validation, and testing
References:
Other resources (click)...
Coursera slides (introductory):
Additional code examples:
Workbook : CSV/TSV/JSON; extracting simple statistics; pandas; plotting
Wk 1/2 Supervised Learning: Classification

Logistic regression
Other Classification Algorithms
How to evaluate classifiers
References:
Other resources (click)...
Coursera slides:
Code examples:
Workbook 2 : Classification; gradient descent
Workbook 3 : Classification diagnostics; training/testing
Wk 3/4/5 Recommender Systems

Collaborative Filtering
Latent Factor Models
Recommender Systems Evaluation
Deep Learning for Recommendation
References:
Other resources (click)...
Coursera slides:
Code examples:
Past midterms (click)...
Sentiment analysis
Bags-of-words
TF-IDF
Stopwords, stemming, and low-dimensional representations of text
References:
Other resources (click)...
Wk 7/8 Content and Structure in Recommender Systems

Factorization Machines
Group- and Socially-Aware Recommendation
Online Advertising
(Maybe some tools/libraries/misc topics, depending on time)
References:
Complementary Item Recommendation
Fashion and Outfit Recommendation
Fit Prediction
References:
9 Modeling Temporal and Sequence Data

Sliding windows and autoregression
Temporal dynamics in recommender systems
References:
Filter Bubbles and Recommendation Diversity
Calibration, Serendipity, and Other "Beyond Accuracy" Measures
Algorithmic Fairness
References: