AI Powered Multiple Disease Detection

Sudiksha Asati

Integrated B.Tech + M.Tech Amity University

India sudiksha.asati@s.amity.edu

Dr. Ekta Soni Asst.Professor Amity University

Dr. Sakshi Kathuria

Asst.Professor Amity University

Abstract—The growing accessibility to digital records of crimes has led to the possibility of data-driven public safety planning, but the current methods of analytic processing do not have much to offer in the way of complex spatio-temporal dependencies, multi-type crime correlations, and operationally-sensitive insights of fairness. The proposed Crime Data Analytics System of Pattern Detection and Decision Support combines deep spatio- temporal learning, graph-based modeling, explainable artificial intelligence and fairness-aware optimization in a single model and optimizes the learning of crime intensity and emergent hotspots.The study develops region- and feature-level attributions in interpretability modules to increase transparency. Fairness audits systems minimize the geographical and socio-economic variations in recommendations of patrols. Multi-source urban datasets, such as Indian deployments consistent with NCRB-style records, are experimentally shown to be better in both forecast- ing accuracy and hotspots coverage than traditional machine- learning baselines. The findings emphasize the possibility of the hybrid, ethically-driven analytics to assist in proactive policing and evidence-based allocation of resources in contemporary cities.Index Terms—Crime Analytics; Spatio-Temporal Prediction; Graph Neural Networks; Hotspot Detection; Decision Support Systems; Explainable AI; Fairness in AI; Predictive Policing; Urban Computing; Public Safety Analytics.