AUTONOMOUS AI AGENTS FOR DATASCIENCE

Abstract

In the modern era of data-driven decision-making, data science has become a crucial field for extracting meaningful insights from large volumes of data. However, traditional data science workflows involve multiple complex and time-consuming steps such as data preprocessing, exploratory data analysis (EDA), model selection, training, and evaluation. These processes require significant human effort, technical expertise, and continuous monitoring, making them challenging for beginners and inefficient for large-scale applications. To address these issues, this project proposes the development of an Autonomous AI Agent for Data Science, which aims to automate the entire data science pipeline. The proposed system is designed to perform key data science tasks with minimal human intervention. It begins by allowing users to upload datasets through a userfriendly interface. The system then automatically processes the data by handling missing values, removing inconsistencies, and preparing it for analysis. Following this, exploratory data analysis is conducted to identify patterns and relationships within the dataset. Based on the nature of the data, the system intelligently selects appropriate machine learning models, trains them, and evaluates their performance using standard metrics. The implementation of this system is carried out using Python and various libraries such as Pandas, NumPy, Matplotlib, and Scikit-learn. Additionally, Streamlit is used to develop an interactive web-based interface that simplifies user interaction and enhances accessibility. The modular architecture of the system ensures flexibility, scalability, and ease of maintenance, allowing it to be extended with additional features in the future.