PM.

Hi, I'm Przemek.

A PhD researcher specializing in data science & analytics.

I use data to find answers, solve problems, and make things more interesting. Data is a reflection of us – our biology, choices, behaviors, and interactions. I like discovering the stories it can tell.

Projects

Prisoner's Dilemma game

Simulation of the the Iterated Prisoner's Dilemma game using various strategies. Agents with different strategies compete in a tournament, and the average score per move for each strategy combination is calculated. The score distribution for each strategy is visualized, providing insights into the performance of different strategies.

  • Python
  • Game Theory
  • Visualization

Two-armed bandit game with reinforcement learning analysis

A Two-Armed Bandit game application where players choose between two arms offering different reward potentials. The players' choices are analyzed using a reinforcement learning model, and the decision history characteristics are visualized. Explores the dynamics of decision-making and learning from experience in a reinforcement learning scenario.

  • R
  • Shiny
  • Reinforcement Learning

Simulation of Schelling's model

Simulating a variation of Schelling's model of segregation. Agents of two types are randomly distributed throughout the grid. Each agent decides if it is happy based on the types of neighboring agents. Unhappy agents move to an empty space on the grid, simulating the dynamics of segregation. Implemented on a toroidal grid with configurable parameters and visual representation of agent distribution and happiness over time.

  • R
  • Simulation
  • Visualization

Simulation of behavior spread in a social network

Simulating the spread of behaviors within a grid-based social network. Agents adopt behaviors based on specific probabilities determined by their neighbors' behaviors. The dynamics of behavior spread are calculated and visualized, demonstrating how behaviors propagate through social networks and the factors influencing their adoption.

  • R
  • Simulation
  • Visualization

Dog growth prediction model and application

Building a Bayesian predictive model for dog growth trajectories based on the von Bertalanffy growth curve. The model is implemented in a Shiny app that generates personalized growth predictions based on user input. Showcases the application of Bayesian modeling techniques to predict biological growth patterns.

  • R
  • Shiny
  • Bayesian Modeling

Simulation of the Game of Life

Simulating variations of Conway's Game of Life, a cellular automaton devised by mathematician John Conway. The Game of Life is a zero-player game where the evolution of the grid is determined by the initial state and a set of rules governing the birth, survival, and death of cells. Explores emergent behavior patterns arising from simple rules in cellular automata.

  • R
  • Simulation
  • Visualization

Maze generation and solving

Generating random mazes using the Recursive Backtracking algorithm and solving them using a chosen path-finding algorithm. The generated mazes and their solutions are visualized, allowing for comparisons of the strengths and weaknesses of the different path finding algorithms.

  • Python
  • Algorithms
  • Visualization

Simulation of forest fire dynamics

Simulating the spread of forest fires using a stochastic grid-based model. The model tracks the states of individual trees and visualizes the progression of fire on a forest landscape. By capturing the dynamic interactions between cells, explores the behavior and patterns of forest fire propagation under different conditions.

  • R
  • Simulation
  • Visualization

Simulation of random walks in one, two, or three dimensions

Simulating random walks in one, two, and three dimensions. Includes R implementations for 1D and 2D simulations, and a Python implementation for the 3D simulation. Agents move randomly within their respective spaces, and their trajectories are visualized. Parameters to control the number of agents, number of steps per agent, and random or centralized starting positions (2D and 3D) are configurable.

  • R
  • Python
  • Simulation
  • Visualization

Skills

Contact

Email me