Manik-Kapil/CoviShield-Predictive-Measures-for-COVID-19-Community-Transmission

IBM Applied Data Science Professional Certificate Capstone Project # Predictive Measures for Mitigation of COVID-19 Community Transmission > **Status:** Archived (Project completed April 2020) > **Author:** Manik Kapil ## 📌 Project Overview This data science project, developed during the critical early phase of the COVID-19 pandemic (April 2020), aims to mitigate community transmission through predictive modelling and geospatial analysis. The system analyzes infection hotspots across India to scientifically recommend optimal locations for **quarantine facilities**, **isolation wards**, and **community kitchens**. By leveraging K-Means clustering on district-level data and integrating Foursquare API location intelligence, the project provides a strategic framework for resource allocation in worst-hit regions. ## 🔍 Exploratory Data Analysis (EDA) & Inferences Extensive analysis was conducted on daily patient databases, time-series case data, and testing metrics to derive the following critical inferences: ### 1. Severity & Regional Hotspots * [cite_start]**High-Load States:** Maharashtra emerged as the epicenter with **8,068 confirmed cases**, followed significantly by Gujarat and Delhi [cite: 74, 78-82]. * [cite_start]**Active Case Load:** Analysis of active cases revealed that while Maharashtra had the highest volume (6,538), states like Delhi (1,702 active cases) showed promising recovery trends, potentially attributed to specific interventions like Plasma Therapy [cite: 172-177, 186-192]. * [cite_start]**Mortality Analysis:** Maharashtra recorded the highest fatalities (342), followed by Gujarat (151) and Madhya Pradesh (103), highlighting the urgent need for critical care infrastructure in these specific zones [cite: 234-250]. ### 2. The "Lockdown Effect" Analysis A core objective was to determine if India's pre-emptive lockdowns successfully delayed "Stage III" (Community Transmission). * [cite_start]**Linear vs. Exponential Growth:** Post-lockdown analysis indicated that while daily confirmed cases rose (partially attrib …