Open Access Open Access  Restricted Access Subscription Access

AI Analyst - AI Powered Data Analytics Platform

Mahashree A M, Monica P

Abstract


The rapid expansion of data-driven applications demands efficient and accessible tools for machine learning model development and business intelligence. This paper presents AI Analyst ULTRA — an interactive web-based data analytics platform that enables end-to-end automation of the data science workflow. The system accepts CSV datasets uploaded by the user, automatically performs data cleaning and preprocessing, delivers rich interactive visualisations, and applies multiple machine learning techniques including Regression, Clustering, and Anomaly Detection. AI-driven insights and business decisions are generated through integration with the Gemini API. The platform is evaluated on benchmark datasets and demonstrates strong analytical capability across diverse domains. The proposed platform lowers the barrier for non-expert users to apply machine learning and provides a reproducible, extensible baseline for automated analytics research.


Full Text:

PDF

References


M. Feurer et al., "Efficient and Robust Automated Machine Learning," NeurIPS, vol. 28, 2015.

C. Thornton et al., "Auto-WEKA: Combined Selection and Hyperparameter Optimization," KDD, 2013.

E. LeDell and S. Poirier, "H2O AutoML: Scalable Automatic Machine Learning," AutoML @ ICML, 2020.

L. Breiman, "Random Forests," Machine Learning, vol. 45, no. 1, pp. 5–32, 2001.

M. Ester et al., "A Density-Based Algorithm for Discovering Clusters," KDD, pp. 226–231, 1996.

V. Chandola et al., "Anomaly Detection: A Survey," ACM Comput. Surv., vol. 41, no. 3, 2009.

X. Cheng et al., "GPT for Data Science: Capabilities and Limitations," arXiv:2305.03269, 2023.

F. Pedregosa et al., "Scikit-learn: Machine Learning in Python," JMLR, vol. 12, pp. 2825–2830, 2011.

W. McKinney, "Data Structures for Statistical Computing in Python," SciPy, 2010.

Streamlit Inc., "Streamlit: The Fastest Way to Build Data Apps," https://streamlit.io, 2024.


Refbacks

  • There are currently no refbacks.