Prioritize-Vaccine-Delivery

EY Hackathon Project

Prioritising Vaccine Delivery πŸ’‰

An AI/ML Project πŸ€–

All Contributors
Forks Stars

EY Header

The Intro πŸ“Œ

This repository is created towards developing a solution for the project entitled Prioritize vaccine delivery using AI/ML of EY Techathon - 2020.

more information about this project can be found here.

Objective πŸ€”

Our objective is to provide a solution to the problem of where the vaccine should be delivered first, accross the whole nation. Furtheremore, it’ll be even better if we could predict the percentage of total dose that should be delivered there.
The priority can be classified on the base of multiple criterion, viz.:

  1. states or union-teritories
  2. districts or provinces
  3. age-group
  4. employees’ group (doctors, police, volenteers etc..)


Requirements for the Desired Solution ⚑

The solution should enable one to get the following:

  1. a visual representation of prioritized states on map for the whole country
  2. a visual representation of prioritized districts on map for every state
  3. States and Districts with highest requirement of the vaccine must always be visible.
  4. Detailed priority (as mentioned in objective section) requirement for any searched state should also be possible.


Steps of the Solution 🐾

We can divide the whole process of this solution into three major parts.

  1. Prioritization for the whole nation
  2. Prioritization for the every individual disrtict
  3. UI intigration for the solution

Elaborating Further

1. Prioritization for India πŸ”

Here, we’ll have to inform that which part or state or population-distribution should get the vaccine, first. For that we have to keep various things into consideration.

States can be classified on the following criterion:

  1. Affect of COVID-19
    • by zone (red, orange and green zone by gove.)
    • by percentage of population affected, out of total
    • by death VS recovery ratio
  2. Population Distribution
    • by age-group
    • by employment status
    • by literacy-rate
  3. Economy & Development
    • Connectivity to other states/district (through transportation etc..)
    • Interaction among the population

2. Prioritization for a Particular State ☝️

For a better estimate and results, it is good idea to provide another prioritization in the state-level. It will be helpful for the state governments to manage the delivery in an optimal way.
This prioritization can be done district-wise.

District can be classified on the following criterion:

  1. Affect of COVID-19
    • by zone (red, orange and green zone by gove.)
    • by percentage of population affected, out of total
    • by death VS recovery ratio
  2. Population Distribution
    • by age-group
    • by rural VS urban ratio
    • Interaction among the population

3. UI Integration πŸ’»

We wish the home screen to have the following structure:

    HOME
    .
    β”œβ”€β”€ Map (India)        # Bubble Chart for Priority (district-wise)
    β”‚
    β”œβ”€β”€ 5-States           # with hishest priority 
    β”‚
    β”œβ”€β”€ 5-Districts        # with hishest priority 
    β”‚
    β”œβ”€β”€ Population-Wise    # with hishest priority 
    β”‚   β”‚
    β”‚   β”œβ”€β”€ Top Age-Group 
    β”‚   β”‚
    β”‚   └── Top Employees
    β”‚
    β”œβ”€β”€ Live COVID-19 Stats   # Overall India 
    β”‚
    └── SEARCH                # Detailed info for any perticular state
        .
        .
        └── ...         

One should be able to discover the information about any perticuler state on searching.
This new page should consist of the following:

    STATE
    .
    β”œβ”€β”€ Map (State)        # Bubble Chart for Priority (district-wise)
    β”‚
    β”œβ”€β”€ 5-Districts        # with hishest priority 
    β”‚
    β”œβ”€β”€ Population-Wise    # with hishest priority 
    β”‚   β”‚
    β”‚   β”œβ”€β”€ Top Age-Group 
    β”‚   β”‚
    β”‚   └── Top Area          # rural or urban
    β”‚
    └── Live COVID-19 Stats   # For current State 


Data-Sources ℹ️

https://www.covid19india.org/
https://api.covid19india.org/


Wanna Countribute? ✨

You can contribute to this project in various way. Few of them can be:

  1. Providing datasets (as explained above)
  2. Python scripts to scrape, get or clean the datasets
  3. Machine Learning models to classify the datasets on the above mentioned criterion
  4. Designing UI-frameworks (Django or Flask will be highly prefered)
  5. Enhencement in documentation