Every country in the world is associated with infinitely many attributes, both negative and positive. With widely varying factors like poverty, economy, literacy rate, life expectancy etc, it becomes difficult to qualitatively determine how well off a country actually is, or will be in the future. This inspired us to create Wellness Forecast.
What it does
Wellness forecast is equipped to predict the wellness index of every country until 2030. It graphically represents this information on a heat map of the world.
How we built it
Firstly, we collected data for every country from the UN world bank for the past sixty years. Then we processed and formatted the data using Pandas. After that we made use of Scikit learn to train a model on the input features, and predicted the values of the features in the next thirteen years. Consequently, we formulated another model to calculate the wellness index based on our features. Using this wellness index, we projected the data on a heat map to effectively convey the wellness of every country in the world. Moreover, we designed an interactive website using HTML5 and CSS3 to display our findings.
Challenges we ran into
One of the most critical challenges we faced as a team was collaborating with one another and making our individual segments of code work seamlessly with one another. Machine learning target values were very erratic, and data was very irregular.
Accomplishments that we're proud of
We were able to process large amount of data in a usable way. We were able to predict reasonable values.
What we learned
Skills in Web Development(CSS), Google Maps API and Machine learning algorithms, Pandas and Sklearn.
What's next for Wellness Forecast
Finding correlation between various features, making maps more interactive and experimenting with larger data sets and different machine learning algorithms.
Team member codes - 455, 451, 452, 450.