Contact - norbert.frydrysiak@proton.me

I approach my studies with great ambition and passion. I am seeking an IT internship that will provide me with valuable experience necessary to achieve my dreams. Despite being an individualist, I know how to function in a team to develop and achieve our goals together. I am open to new challenges.

From a very young age, I have shown a great interest in the exact sciences, especially mathematics. I am also interested in the family of GNU/Linux operating systems and new technologies. In my free time, I follow the world of football and computer games. My newest interest is Reinforcement Learning, and it's something I'm eager to further develop in the future.

Tools and Technologies

I am going to learn

• Bachelor’s Thesis Repository: Adapting the gips library for classification problem utilizing discriminant analysis – gipsDA
  Repository containing my Bachelor’s thesis project, including the thesis document, manual tests, and datasets used for research and analysis.

• Bachelor’s Thesis Package: gipsDA
  Implementation of the software package developed as part of my Bachelor’s thesis, along with unit tests and documentation.

• Transformers Speech - This project investigates the performance and optimization of deep learning architectures for keyword spotting on the TensorFlow Speech Commands dataset, comparing custom CNN and Audio Transformer models trained from scratch against transfer learning architectures like Wav2Vec 2.0, and exploring audio augmentation and hierarchical cascade classification to handle challenging minority classes.
• CNN Image Classification - This project investigates the performance and optimization of Convolutional Neural Networks for image classification on the CINIC-10 dataset , comparing models trained from scratch against transfer learning architectures and exploring few-shot learning paradigms.
• Multi Source Market Analytics - This project implements a comprehensive Big Data architecture designed to ingest, unify, and analyze market data streams from various heterogeneous exchange APIs.
• Computational Intelligence - This project contains from-scratch implementations of Genetic Algorithms, Neural Networks, and Self-Organizing Maps in Python to solve a variety of practical optimization and data analysis problems.
• Financial Math Intro - This project explores computational financial mathematics by implementing Monte Carlo methods for option pricing within the Black-Scholes and Cox-Ross-Rubinstein models.
• WB-SAE-CBM - This project explores Sparse Autoencoders (SAEs) as concept bottleneck models (CBMs) for interpreting learned representations in vision models.
• Top5 Football Analytics - A data warehouse project enabling comprehensive analysis of football players and teams from Europe’s top 5 leagues using scraped data from FBref, Transfermarkt, and Wikipedia.
• Bayesian Image Restoration - Implementation of Bayesian statistical techniques for restoring degraded images through MAP and MMS estimators, utilizing Potts model and truncated quadratic energy function.
• March Machine Learning Madness 2025 - Kaggle Tournament - Predicting the outcomes of every possible matchup in the 2025 NCAA men's and women's basketball tournaments using machine learning.
• AUTOFUND - The project is an AutoML package that automates building, training, and tuning machine learning models on Numerai data for market trend forecasting.
• Hyperparameter Tunability AutoML - The goal of this project is to analyze the tunability of hyperparameters of three selected machine learning algorithms (e.g., XGBoost, Random Forest, Elastic Net) on at least four datasets.
• Django-Blog-Forum - This project is a personal website functioning as a blog and forum, developed using Django, Python, and HTML.
• RL-Doom - The Project implements Reinforcement Learning methods in Doom game using vizdoom package.
• Brand Laptops Clustering - The objective of the project is to analyze and evaluate various clustering algorithms on a dataset of laptops.
• Credit Score Classification - The goal of the project is to predict based on data whether a given person will repay the loan.
• Linux Me Project - Analyzing Linux system data using R and Shinydashboard for insightful insights.
• Java Football Data Visualizer - Football FBref Data Scraper with Plotly Visualization in Java.
• Fast Food Data Analysis Project - The goal of the project was to analyze food-related data using R language and ggplot package, and create a poster.

⚡ Github Stats