Agentic AI Swarm Optimization using Artificial Bee Colonization (ABC)

This project implements the Artificial Bee Colony (ABC) swarm intelligence algorithm using an Agentic AI framework for adaptive grid search. The system uses Google's Gemini LLM to power three types of autonomous bee agents—Scout, Employed, and Onlooker—whose roles are defined through natural language prompts. Applied to a clustering problem on the Iris dataset, these agents explore, refine, and select optimal hyperparameters for scikit-learn clustering algorithms. The collective goal is to improve clustering performance, with a fitness model evaluating solutions based on the Adjusted Rand Index (ARI). This approach demonstrates a creative integration of LLMs into a classic data science optimization workflow.