A large portion of the class grade will be the work you upload to your portfolio. This will be a combination of code, code narrative, reports, and presentations.
In my public repository, ml-portfolio you will find the code and notes I write while learning in Dr. Karen Mazidi's class Introduction to Machine Learning at the University of Texas at Dallas. While this is currently just a markdown document with links to posted materials, I hope to expand this in the future, perhaps to hold resources accessed by a live site.
- Finish setting up the main site at zaiquiriw.github.io, goals can be found here
- Port theme and set up Mkdocs
- Set up Mkdocs build action
- Must use absolute file paths until I change because I do not want to touch a jekyll config file for fixing relative file paths
- Decide whether to build site in
gh-pagesbranch or in/(root) - Host resources both as navigable html files, and link to pdfs.
- I want to figure out how to easily format a download link in a markdown page, after the static generator has converted it to html. Saving this!
- Set up Mkdocs build action
- Create a GitHub portfolio for class work
- Summarize the main branches of ML algorithms
- Reflect on your personal interest in ML
An Overview of ML can be found in the repo! You can read it on here as well
- Read a CSV file in C++
- Create a suite of simple stats functions in C++
- Find more complex values:
- Covariance
- Correlation
- Write documnetation on the value of this data
As we are learning R, we need a refresher on C++, so this was that. The documentation explains the importance of what I went over, and the code practice can be found here.
Well this was hard! Work on linear regression can be found here and linear classification can be found here!
Instead of just messing around with logistic regression and naive bayes, we build it from scratch. A blurb about the work can be found here
Group projects are hard! but we got it [here]("https://github.com/zaiquiriw/ml-portfolio/blob/main/similarities/ML Similarities Narrative.pdf") Code
- https://github.com/zaiquiriw/ml-portfolio/blob/main/similarties/Clustering.pdf
- https://github.com/zaiquiriw/ml-portfolio/blob/main/similarties/Dimensionality.pdf
- https://github.com/zaiquiriw/ml-portfolio/blob/main/similarties/Regression.pdf
- https://github.com/zaiquiriw/ml-portfolio/blob/main/similarties/Classification.pdf
We iterate further down the ml rabbit whole with improved classification and regression methods, using some fun math and the art of doing things over and over. The algorithms used are noted here, and implemented in regression, classification, and various ensemble methods.
Just ran through a couple use cases of python for ml, read here
This has been a fun semester of ML! With my resulting understanding of previous ML concepts, its nice to know I understand what improvements my neural networks shown here have over earlier and simpler classification models.