AI Agents

This project involved designing and implementing autonomous intelligent agents, progressing from simple reactive models to complex deliberative architectures. It explored symbolic, behavioral, and hybrid paradigms of artificial intelligence, integrating state machines, behavior arbitration, search algorithms, and Markov Decision Processes to enable agents to plan, adapt, and optimize their actions in dynamic environments. The practical evolution demonstrates how combining different AI paradigms results in robust agents capable of immediate reactions and strategic long-term planning.

Image Processing

This project explores fundamental concepts in multimedia information processing, including image analysis, compression, and error control coding. It applies techniques such as color space conversion, histogram analysis, and bit-plane decomposition to understand image properties and uses quality metrics like SNR and PSNR to evaluate compression results. The work also simulates data transmission over noisy channels, implementing and comparing repetition and Hamming codes to demonstrate their effectiveness in improving image reliability.

Probabilistic Modeling

This project applies Bayesian inference and information theory to solve practical problems, using Python for statistical modeling and simulation. It demonstrates how to estimate parameters for a biased coin using posterior distributions and how to perform Bayesian classification to determine the most probable data compression algorithm. The work also explores the concept of entropy in discrete random processes, validating theoretical results through large-scale simulation.