Inspiration

As Covid-19 cases increases, it brings a lot of misinformation and uncertainty. After the launch of vaccines there are many rumors that it is not that effective and many other information those are misleading to people. To prevent this we thought of working towards the correlations of Covid-19 cases and its vaccines and their trends, so it will help people to understand its behavior and mitigate the uncertainty.

What it does

The project will help you understand the behavior of Covid-19 Cases and Vaccines trends and will help you forecast the situation based on the covid-19 vaccines series completion and other vaccines parameters. It will show different trends and comparison between different dates and time and locations as well.

How we built it

We have used two different datasets from CDC website, which URL are shared in the presentation. At first we did some data wrangling for visualization purpose so it will help understand and make it readable. We have used Seaborn library for visualization as it have very interactive and mature charts and library. Further we have used Python, Pandas and NumPy to wrangle and sample our datasets and bring it in a form so that we can train our dataset.

We have used KNN algorithm to train our dataset using 10 cross fold validation and achieve up to 75% accuracy. We have introduced different classification attributes in the dataset so that system will classify if the covid-19 cases in the upcoming month will increase or decrease.

Challenges we ran into

As the dataset was too big, it took sometime to download and we have used Colab notebook for our implementation that was a free version so that every time we load a file, it removes it if the session is out so we have to upload it every time.

Accomplishments that we're proud of

We accomplished to find the relation between two dataset

What we learned

We have learned how we can sample our dataset and understand the different behavior of dataset while doing data wrangling and preparing dataset for machine learning model.

What's next for CGI EDA - Impact of Covid-19 Vaccines on Covid-19 Cases

We can make it more mature and solid, if we create a secure public mobile app or web app and ask people to submit their feedback about their situation and allow our system to understand and train the data on runtime so it will be available to everyone. Plus we can show the visualization to the apps so that everyone can take advantage of the app.

Also we can help make the project more mature, if we introduce other dataset that has an impact over Covid-19 cases like weather condition, medical conditions etc.

Built With

Share this project:

Updates