Back to Projects

Team Name:

Hackerknights


Team Members:


Evidence of Work

Covidvaccine-ready

Project Info

Team Name


Hackerknights


Team Members


Naga Aditya Kumar G , Nazmul Kaonine , Afrida , Jan , Quang , Mahesh

Project Description


Imagine, it’s December 2020. There are long winded queues in front of the pharmacies and vaccine distribution centres. The Australian government only has a very limited number of vaccines. The population is desperate to return to normality and the vaccine is the only key to achieving this. For the government to reach the minimum requirement herd immunity needs to be reached. With a population of 25 million and barely 15 million vaccines. How will the government succeed in effectively distributing the vaccines?

Fortunately, a small team called the Hacker-Knights come to the rescue.

A novel way of achieving herd immunity and a simple prediction tool with intuitive visualisation filters, that will assist the government in taking control of COVID-19 in the most efficient manner.

The idea is to achieve below outcomes with the government data:
1. Forecast for the covid impact for future months with the existing covid factors (e.g: cases on daily basis, confirm, recovered, deaths and infection rate etc.)
2. Predict the minimum number of vaccines required to achieve herd immunity of a particular region. We considered R-nought, a variable factor, as 2.2 (average according to available sources) hypothetically for the limitation of data and time constraint.
3. Find out the optimal way of distributing available vaccines to the most affected regions based on multiple factors like business impacts, available distributed channels, social, economic conditions and also to achieve herd immunity across the region as fast as possible.


#covid19 government vaccination public epidemic

Data Story


The datasets used come from various sources including Australian Bureau of Statistics, Github open source data and Australia Population (2020) Worldometer. One of the major challenges has been to extract precisely the most relevant data by combining different unstructured datasets into a single effective aggregated dataset that has been normalised. Having the monthly statewise covid cases, deaths and population we extrapolated the number of vaccines needed for each region and improved the prediction efficiency using factors such as employment and unemployment. An interactive visualisation web dashboard(shown in the video) has been built for users.

This model takes into consideration covid cases that have been confirmed and deaths per state over a time period to calculate the R naught value. These cases have been forecasted upto march 21 using multiple linear regression.

After calculating R naught found to be 2.5 on average. We calculated the number of vaccines needed to reach herd immunity for each state according to its population. Formula for Herd immunity: 1/(1-Rnaught) We have run this algorithm to our forecasted range of Mar 21.
If we get limited number of vaccines, the algorithm calculates the optimal distribution of vaccines to take smaller steps towards herd immunity (no more a pandemic).

Future implementation could use more features(density, social distance, house rents, food habits) to train a Neural Network model for the weighted and variability factor to achieve the real time R-naught more accurately to eradicate COVID-19.


Evidence of Work

Video

Homepage

Team DataSets

Australia employed and unemployed stats

Description of Use To predict no.of vaccines and understand the impact on distribution

Data Set

Australia Demographics data and population rate

Description of Use To predict on the no.of vaccines based on the Population

Data Set

Australia Demographics data

Description of Use To predict on the no.of vaccines based on the Population

Data Set

Covid 19 cases for Australia

Description of Use This data has been used to analyse the forecast for future months

Data Set

Challenge Entries

Identifying government investment for enhancing community safety

How can governments apply resources so as to proactively address social issues before they escalate and require reactive interventions (e.g. social work, law enforcement etc.)?

Go to Challenge | 16 teams have entered this challenge.

Improving government services with open data

How can we improve government services by using open government data?

Eligibility: Create a data story in any format demonstrating how government services can be improved by making data-driven decisions. Teams can use any open datasets, but at least one must be a NSW Government dataset.

Go to Challenge | 14 teams have entered this challenge.

COVID-19 Economic / Social Recovery

How might we use data to assist our understanding of what a post-COVID-19 Victoria may look like?

Eligibility: Participants must use one or more datasets from data.vic.

Go to Challenge | 8 teams have entered this challenge.