One of the most famous modern machine learning training datasets is to predict survival of passengers on the titanic. I used this project while experimenting with KNearestNeighbor classifiers, SVMs, LogReg, pipelines (and the problems with dummying categorical data in pandas), as well as the TFlearn front-end for Tensorflow. My sourcecode for that that project can be found here.
Anime_Rec is a Data Science project to generate Anime recommendations based on publicly available data from the website myanimelist.net. The best current writeup(and source code) can be found here. That write-up will be expanded (and broken into smaller parts) in the following way: Part1: Overview (Architecture, Design, Purpose) Part2: Data Collection, EDA. Part3: Generating possible recommendations (Cosine Similarity methods) Part4: Predicting user scores(With Neural Nets) Part5: Productionizing this architecture.
Lol_Scout was a Data Science/ Esports analytics project I pursued to answer questions about how League of Legends players champion choices can be optimized. The current soucecode and readme can be found here. Part 1: Background, Initial Design, Data Gathering, data acquisition lessons learned. Part 2: Modelling with Neural Net Classifiers, XGBoost, conclusions, ways to pursue the project further in the future.
After some time in dormancy, it is once again time for Drunken Thieves to be active. While originally a launchpad for some creative side projects (Film and Games), it became a blog that would go on to become insidethekraken. As one of the founding members of Drunken Thieves, I’ve decided that like me, it’s time for this site to transition. Going forward, this site will be a portfolio/blog about my current focus on Data Science, Machine Learning, and DevOps. While that is my current focus, I am still available for ACSR or CCNA consulting: I’ve […]
So, Emmy Live Blogs aside, you have probably noticed a bit of a lull in the blog for the past month or so. I announced a couple of months ago that some big changes were coming soon, and that time is now. Starting Monday, the Drunken Thieves blog will be no more, as we transition to a new blog. Drunken Thieves has been an incredibly positive experience for Kyu, Keskel, and I, but we all felt it was time for a change. When starting this blog, it was really more for us than for any potential readers, […]