The challenge sounded interesting so we decided to try to solve it.
What it does
It's a trading bot which makes it's decisions based on historic data.
How we built it
We tried to predict the future prices with three different solutions.
Challenges we ran into
The model which was based on ARIMA was not precise enough. This whole area was new for all the team members this caused difficulties.
What we learned
We learned a lot about time series analysis, ML, statistics.