Showcases of Different ML Algorithms

Using Background Color Classification

This is a Kotlin application that experiments with different machine learning algorithms to recommend a light/dark font against different background colors.

Some algorithms are built completely from scratch but others are showcased using a library.

The current algorithms/library implementations:

• Formulaic
• Linear regression (w/ hill climbing)
• Logistic regression (w/ hill climbing)
• Decision Tree
• Random Forest
• Neural Network (w/ hill climbing)
• Neural Network (w/ simulated annealing)
• OjAlgo Neural Network
• DL4J Neural Network

Planned to be added:

• Linear regression (w/ gradient descent)
• Logistic regression (w/ gradient descent)
• Continuous Naive Bayes