Joel Dirisam

I'm a software engineer. I build full-stack web apps with Python and React, and I like explaining how machine learning works.

A logistic regression model learning to separate two groups of points.

Projects

Logistic regression on DonorsChoose.org
Predicts which project proposals on DonorsChoose.org are likely to be accepted, tuned to maximize the area under the curve.
Dimensionality reduction on Amazon food reviews
Applies t-SNE (t-distributed stochastic neighbor embedding) to Amazon food review text.
Wine quality prediction
Uses SVM and k-nearest neighbors, and finds which features help and which lower the model's accuracy.
Exploratory data analysis
Explores datasets to draw out insights and understand the data before modeling.
Stay Connected
A Windows app that tells you when friends are within a chosen radius. Built in C# in 2014 for Microsoft's Windows App Development Program.

How logistic regression works
A simple algorithm that often beats more complicated ones, and a walk through how it works.
Why do we need to standardize the data?
A practical explanation of standardization, worked through with three data points.

I'm a husband, a father and a software engineer.