All docs related to caojoshua's self guided courses. Includes things I've completed, and a bunch of things I may work on the future. The list is large and may get larger overtime, so I will probably never get to most things.
I tend to favor courses that have many programming related assignments, and not as much handwritten assignments. Courses that offer slides or video lectures are both fine. Both undergraduate and graduate courses are suitable. For topics that I just want a quick intro to, I would lean towards undergraduate/quarter system courses. If I want more in-detph knowledge, than I prefer graduate/semester system courses.
- MIT 6.858 Computer Systems Security
- Stanford CS144 Intro to Computer Networking
- Stanford CS 140 Operating Systems
- MIT 6.820 Fundamentals of Program Analysis.
- There are a lot of other program analysis class, but this one is included in MIT OCW and the material is the most accessible
- Did the first assignment. Find the content hard with lack of video lecture, and slides are too barebone. Putting it on halt. Maybe try alternatives instead?
- About half of class is on type theory
- CMU 355 Program Analysis
- alternative to MIT program analysis, although content is a little different. nothing on type theory.
- material is potentially better for self learning. quick glance shows they have detailed notes.
- CMU also offer a grad class but latest public offering is from 10ish years ago and they don't even publically release resources
- Stanford CS 243 Program Analysis and Optimization
- Includes main compiler optimizations and alias analysis
- About half of labs are programming, and the other half are handwritten problems and proofs
- Stanford CS 140E
- labs
- see instructors github for diffent offerings of the same coures
- similar to CS 140. Implement main OS things like threads, virtual memory file systems. Also include some things I don't understand so well yet such as interrupts, bootloader, device drivers
- all runs on raspberry pi. would need to purchase one myself.
- Based on instructor estimate, should take around 60 hours total to complete
- later labs are not documented, probably because they never got to them. For example, the user processes lab is not documented.
- Stanford CS 140E alternate offering
- relatively more driver related things
- includes most of the main things as the other offering
- seems virtual memory was cut off from last lab. they probably ran out of time.
- assignments in rust
- would probably prefer this offering
- Stanford Parallel computing
- Not sure if this subject requires taking a course, but it is a challenge of computing today so maybe its good to learn more about it
- They claim to be implementation heavy
- Some labs require special hardware ie. intel with SIMD instruction support, and nvidia GPU with CUDA support
- MIT Network and Computer Security
- Databases
- Stanford CS 231n seems like a good course for ML noobies, and labs are implementation based
- Should just get better at algorithms in general, cause I'm bad. Could be a course or just read books.
- Domain specific algs ie. crypto, optimizations, formal languages
- graphics
- Math: Differential Equations, Statistics, Abstract Algebra, combinatorics
- MIT open course ware
- Stanford course catalog
- Berkeley course catalog
- Carnegie Mellon course catalog
- Illinois course catalog
- Can't find course catalogs for some top CS schools, but this is enough
- could maybe someday take a look at non-university courses ie. coursera, various cloud trainings