This document desribes/compares some related courses.
There is overlap between CPSC 330 and CPSC 340. However, CPSC 330 is about using machine learning whereas CPSC 340 involves implementing machine learning algorithms from scratch. CPSC 330 is broader, and includes topics like data cleaning and communicating your results; CPSC 340 goes deeper into the algorithms and the mathematical / numerical considerations underlying them. CPSC 340 has many more prerequisites, especially in terms of math courses. The best choice will depend on your interests and goals. It is also perfectly reasonable to take both courses!
Topics that are only in CPSC 330:
- build an ML pipeline
- data cleaning
- time series data
- communicating your results
- and more
Topics that are only in CPSC 340:
- probabilistic models
- optimization and loss functions (this is a major topic in CPSC 340)
- dimensionality reduction
- and more
Topics that appear in both courses (but typically in a very different way):
- basic classification and regression methods: decision trees, KNN, etc.
- fundamentals of learning
- clustering
- recommender systems
- deep learning
Our department also offers an advanced ML course, CPSC 440, which is approximately equivalent to the graduate course CPSC 540. Compared with this course, CPSC 440/540 has...
- More advanced material.
- More focus on theory/implementation, less focus on applications.
- More prerequisites and higher workload.
For almost all students who haven't yet taken CPSC 340, this is the right class to take.
- CPSC 340 focuses on the most widely-used methods in practice.
- CPSC 440/540 is intended as a continuation of CPSC 340.
- You'll miss important topics if you skip CPSC 340.
You may also want to consider taking related courses from statistics: STAT 305, STAT 306, STAT 406, STAT 460, STAT 461 (as well as EOSC 510). A discussion of the difference between CPSC 340 and these various STAT classes written by a former student (Geoff Roeder) is available here.