mansiarora1009 / analytics

Analysis of the municipal data for real time alerts, predictive analytics and more...

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

This project covers the analysis of the municipal data for real time alerts, predictive analytics and more... We will analyse the data created by the eGov SmartCity ERP for increasing the efficiency of Municial Employees by enabling data driven governance.

DataKind

DataKind is a nonprofit organization that brings leading data scientists together with high impact social organizations through a comprehensive, collaborative approach that leads to shared insights, greater understanding and positive action through data in the service of humanity.

eGovernments Foundation

eGovernments Foundation transforms urban governance with the use of scalable and replicable technology solutions that enable efficient and effective municipal operations, better decision making, and contact-less urban service delivery. eGovernments is looking forward for the following outputs from the engagement:

  • Monthly forecasts for overall top 5 complaint types across the city at the ward level : This will be used by the Commissioner and City Level Municipal Executives to allocate resources and take proactive action.
  • Monthly forecasts for top 3 complaint types for each ward: These will be used by the zonal and ward level executives to take appropriate actions in their respective jurisdictions.
  • Alerts and Notifications system which raises an alert if an anomaly or increasing trend is detected.

Project milestones

Part 1: Data Analysis Report

Analyze historical data to find out what patterns exist in the data. For example, Water Logging complaints may peak in September in wards x, y and z. This report will contain the following details:

  • Ward Level:Cyclical or Seasonal Patterns
  • Complaint Type: Cyclical or Seasonal Patterns, Wards in which these complaints occur the most
Part 2: Top Complaints - Forecast Models (1)
  • A monthly forecast model, which forecasts the number of ward-wise complaints for the top 5 complaint types for the next 3 months
  • The top 5 complaints are picked according to the counts at the city level
  • This model may return a value - ‘Unknown’ for wards which don’t satisfy the assumptions listed below
Part 3: Ward Level - Forecast Models (2)
  • For each ward, forecasts the number of complaints for the top 3 complaint types for the next 3 months
  • The top 3 complaint types are picked according to the counts, at the ward level
Part 4: Alerts and Notifications
  • Alert if an anomaly or increasing trend is detected from the existing predictions
  • Note that this is not fully specified, and would require exploration from the team

How to run

About

Analysis of the municipal data for real time alerts, predictive analytics and more...

License:MIT License


Languages

Language:Jupyter Notebook 60.2%Language:Python 29.1%Language:R 10.7%