Skip to content
View staifmatej's full-sized avatar
0 followers · 1 following
  • Prague Metropolitan Area, Czechia
  • 01:49 (UTC +01:00)
  • LinkedIn in/matej-staif

Block or report staifmatej

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
staifmatej/README.md

Pet Projects

prg-precipitation-forecast-hmm (May 2025, developed over 16 weeks)🐍

  • Implements three Hidden Markov Model variants (Discrete, Gaussian Mixture, Variational) for Prague precipitation forecasting using 25 years of meteorological data. Achieved 64.91% accuracy with GMM-HMM, outperforming naive baseline through Bayesian optimization and sliding window backtesting.

linear-least-squares-methods (July 2025, 2 weeks intensively)🐍

  • Implements four regression models from scratch using three computational approaches (Pure Python, NumPy, Numba JIT). Features performance benchmarking across implementation strategies, curve fitting for sixteen mathematical functions, and demonstrates significant speed improvements with JIT compilation over pure Python.

mortgage_approval_bayesian_network (August 2025, developed over 4 weeks)🐍

  • Implements a Linear Gaussian Bayesian Network using pgmpy for mortgage approval decisions, trained on synthetic historical client data. Features comprehensive stress-testing under economic downturn scenarios demonstrating model robustness and generalization capability. Includes evaluation framework with confusion matrices and ROC curves for risk assessment validation.

👤 Author

My name is Matej Staif, born in 2003 in the north of the Czech Republic. I currently live in the Prague Metropolitan Area in Czechia.

I am currently pursuing a bachelor's degree @ CTU FIT.

If you have any questions, feedback, or find any issues in my repositories, please feel free to contact me via email: staifmat@fit.cvut.cz

Pinned Loading

  1. prg-precipitation-forecast-hmm prg-precipitation-forecast-hmm Public

    This repository contains the implementation of Hidden Markov Models for daily precipitation forecasting in Prague, using 25 years of meteorological data from Prague-Ruzyně station.

    Jupyter Notebook

  2. linear-least-squares-methods linear-least-squares-methods Public

    This repository implements four regression methods (Linear, Ridge, Lasso, Elastic Net) from scratch using three computational approaches with performance benchmarking, fitting and visualization.

    Python

  3. mortgage_approval_bayesian_network mortgage_approval_bayesian_network Public

    This repository implements Linear Gaussian Bayesian Network (LGBN) for mortgage approval decisions using pgmpy, featuring 27-node architecture with synthetic data and stress-testing.

    Python