This webpage is for an old version of the course; content may be out of date!

CSE 158/258: Web Mining and Recommender Systems

Autumn 2020, Monday/Wednesday 17:00-18:20 PST, Twitch

	Basic Info

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.

The course meets twice a week on Monday/Wednesday evenings, starting October 5. Meetings are livestreamed on twitch, but recordings will also be made available here.

There is no textbook for the course, though chapter references will be 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 material.

Office hours:

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

Assessment:

• Homework 1: due Oct 19 (Monday week 3)
• Homework 2: due Nov 2 (Monday week 5)
• Midterm: Nov 9 (take-home: released 6:30pm Monday Nov 9, due 6:30pm Tuesday Nov 10)
• Homework 3: due Nov 16 (Monday week 7)
• Assignment 1: due Nov 23 (Monday week 8)
• Homework 4: due Nov 30 (Monday week 9)
• Assignment 2: due Dec 7 (Monday week 10)

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 26%.
• Each Assignment is worth 25%.
• Assignment 2 is a group assignment. All other assessment must be completed individually.
• All assessments are due before the Monday lecture on the due date. Late submissions are not accepted.

	piazza page (CSE258)

	piazza page (CSE158)

	last year's course webpage

	intro and course outline slides

1	Supervised Learning: Regression

Monday October 5 / Wednesday October 7:

• Least-squares regression
• Overfitting and regularization
• Training, validation, and testing

Other resources:

• Bishop ch.3
• Elkan ch.3,6
• Instructions to access videos on coursera

Coursera slides (introductory):

• CSV and JSON files
• Reading CSV and JSON into Python
• Processing structured data in Python
• Extracting simple statistics from datasets
• Data filtering and cleaning
• Text and string processing in Python
• Time and date data
• Matrix processing and numpy
• Regression in Python
• Features from categorical data
• Features from temporal data
• Feature transformations
• Missing values
• Motivation behind the MSE
• Over and underfitting
• Setting up a codebase for evaluation and validation
• Evaluating a regularized model
• Evaluating classifiers for ranking
• Introduction to Training and Testing
• Validation
• Implementing a regularization pipeline in Python
• Guidelines on the implementation of predictive pipelines

Code examples:

• Workbook 1: CSV/TSV/JSON; extracting simple statistics; pandas; plotting
• Notebook from lecture

Files	week1.py	50k beer reviews	non-alcoholic beer reviews

Lecture materials	lecture 1 video	lecture 2 video	slides	+ annotations

2	Supervised Learning: Classification

Monday October 12 / Wednesday October 14:

• Logistic regression
• SVMs
• Multiclass and multilabel classification
• How to evaluate classifiers

Other resources:

• Bishop ch.4
• Elkan ch.5,8
• More detailed derivation of the SVM (2018)
• Case study: reddit popularity

Coursera slides:

• Nearest Neighbor
• Gradient descent
• Gradient descent in Python
• Classification in Python

Code examples:

• Workbook 2: Classification; gradient descent
• Workbook 3: Classification diagnostics; training/testing
• Notebook from lecture

Files	week2.py	50k book descriptions	5k book cover images

Lecture materials	lecture 3 video	lecture 4 video	slides	+ annotations

3	Dimensionality Reduction and Clustering

Monday October 19 / Wednesday October 21:

• Principal Component Analysis
• K-means & hierarchical clustering
• Community detection

Other resources:

• Bishop ch.9
• Elkan ch.13
• More detailed derivation of PCA (2018)

Code examples:

• Notebook from lecture

Files	week3.py	facebook ego network

Lecture materials	lecture 5 video	lecture 6 video	slides	+ annotations

4	Recommender Systems

Monday October 26 / Wednesday October 28:

• Collaborative Filtering
• Latent Factor Models

Other resources:

• Elkan ch.11

Coursera slides:

• Implementing a similarity-based recommender
• Using a similarity-based recommender for rating prediction

Code examples:

• Workbook 4: Recommender systems
• RSTensorflow: Recommender systems in Tensorflow
• Notebook from lecture

Kaggle pages (Assignment 1):

• Play prediction
• Category prediction (158 only)
• Time played prediction (258 only)

Files	week4.py

Lecture materials	lecture 7 video	lecture 8 video	slides	+ annotations

5	Text Mining

Monday November 2 / Wednesday November 4:

• Sentiment analysis
• Bags-of-words
• TF-IDF
• Stopwords, stemming, and low-dimensional representations of text

Other resources:

• Elkan ch.12
• Recommender Systems Datasets

Code examples:

• Notebook from lecture

Files	week5.py

Lecture materials	lecture 9 video	lecture 10 video	slides	+ annotations

6	(Take-home) Midterm

Midterm	Nov 9

• Midterm released 6:30pm Monday Nov 9
• Midterm due on gradescope 6:30pm Tuesday Nov 10

	sp15 midterm (CSE190)	Solutions	Solution video (starts at 49:55)

	fa15 midterm (CSE190)	Solutions	Solution video (starts at 35:10)

	fa15 midterm (CSE255)	Solutions	Solution video (starts at 32:25)

	wi17 midterm (CSE158)	Solutions	Solution video (starts at 42:00)

	wi17 midterm (CSE258)	Solutions	Solution video (starts at 46:00)

	fa17 midterm (CSE158)	Solutions	Solution video (starts at 35:50)

	fa17 midterm (CSE258)	Solutions	Solution video (starts at 40:15)

	fa18 midterm (CSE158)	Solutions	Solution video (starts at 55:50)

	fa18 midterm (CSE258)	Solutions	Solution video (starts at 45:00)

	fa19 midterm (CSE158)

	fa19 midterm (CSE258)	Solutions	Solution video

Lecture materials	lecture 11 video

6	Tools and Libraries

No lecture	November 11 (Veteran's Day)

Monday November 9:

• Crawling and parsing data from the Web
• Manipulating time and date data
• Simple plotting with Matplotlib
• General-purpose gradient descent in Tensorflow

Code examples:

• Notebook from lecture

Files	week6.py

Lecture materials	slides	+ annotations	lecture 12 video

Assignment	Assignment 2 (due December 7)	slides

7	Data Mining in Social Networks

Monday November 16 / Wednesday November 18

• Power-laws and small-worlds
• Random graph models
• Triads and weak ties
• HITS and PageRank

Other resources:

• Elkan ch.14
• Networks, Crowds, and Markets (book)

Lecture materials	slides	+ annotations	lecture 13 video	lecture 14 video

8	State-of-the-art Recommender Systems

No lecture	November 25 (Thanksgiving)	thanksgiving concert

Monday November 23

• State-of-the-art Recommender Systems
• Bayesian Personalized Ranking
• Factorizing Personalized Markov Chains for Next-Basket Recommendation
• Personalized Ranking Metric Embedding for Next New POI Recommendation
• Real-world Applications
• Recommending product sizes to customers
• Playlist prediction via Metric Embedding

Lecture materials	slides	+ annotations	lecture 15 video

9	Online Advertising

Monday November 30:

• Matching & marriage problems

Wednesday December 2:

• AdWords
• Bandit algorithms

Lecture materials	slides	+ annotations	lecture 16 video

10	Modeling Temporal and Sequence Data

Monday December 7 / Wednesday December 9

• Sliding windows and autoregression
• Temporal dynamics in recommender systems
• Temporal dynamics in text and social networks

Code examples:

• Notebook from lecture

Files	week10.py

Lecture materials	slides	+ annotations	lecture 17 video	lecture 18 video