We are passionate to solve the data analysis challenge!
What it does
With the 64 input attributes (i.e.features), we can predict if a company will go bankrupt.
How we built it
We used various machine learning models to fit the model and get predictions.
Challenges we ran into
Extremely unbalanced dateset! High dimension input data makes training slow.
Accomplishments that we're proud of
We obtained the highest F1_score so far on the leader board
What we learned
Data mining, Machine learning, and visualisation
What's next for Quantum Black Challenge
Improve the model to get more accurate predictions; introduce more thorough predictions such as the probability of a company going bankrupt within a given amount of time