Robin P.M. Kras
LinkedIn
Email
GitHub
Resume
Hello, I am a computer scientist with a specialty in data science and AI.
My technical skills include deep learning, machine learning, data mining, text mining, and all relevant related topics.

Bachelor Thesis Project

Link

Purpose: Exploring Machine Learning Techniques for Classification
Author: R.P.M. Kras | Vrije Universiteit Amsterdam

This thesis evaluates the performance of three popular machine learning algorithms—Random Forest, Support Vector Machine (SVM), and Neural Network—across diverse classification tasks.

Key insights include:

  • Random Forest: Exhibited strong stability and robust performance with categorical and numerical data but struggled with imbalanced datasets.
  • SVM: Delivered the best overall accuracy and reliability for smaller datasets, particularly in multi-class classification.
  • Neural Network: Demonstrated potential in predictive tasks but faced challenges due to dataset biases, underscoring its need for specialized datasets or alternative techniques.
This research emphasizes the importance of dataset selection and model optimization, recommending future exploration of unbiased, larger datasets and tailored techniques for more reliable results.

Keywords: Machine Learning, Classification, Random Forest, SVM, Neural Network

Model 1: Random Forest
Model 2: Neural Network

Masters Thesis Project

To do: Spring 2024.

Additional

Heart disease classification
Predicting $SOL price