# ML4T

This is my solution to the ML4T course exercises. The main page for the course
is [here](http://quantsoftware.gatech.edu/Machine_Learning_for_Trading_Course).
The page contains a link to the
[assignments](http://quantsoftware.gatech.edu/CS7646_Spring_2020#Projects:_73.25).
There are eight projects in total. The summer 2020 page is
[here](http://lucylabs.gatech.edu/ml4t/summer2020/).

To set up the environment I have installed the following packages on my Linux
Manjaro based system.

```
sudo pacman -S python-pandas --asdeps python-pandas-datareader python-numexpr \
               python-bottleneck python-jinja python-scipy python-matplotlib  \
               python-numpy
```

I am also using [mplfinance](https://github.com/matplotlib/mplfinance) to plot
candlestick-charts. You can install mplfinance via pip and find the tutorial
[here](https://github.com/matplotlib/mplfinance#tutorials).

```
pip install mplfinance --user
```

I have included the archived version of the exercise. To extract them run the
following command. The `-n` flag makes unzip never overwrite existing files.

```
unzip -n zips/*.zip -d ./
```

# Reports

- [Report 1](./martingale/martingale.md)
- [Report 2](./optimize_something/optimize_something.md)
- [Report 3](./assess_learners/assess_learners.md)
- No reports for projects 4 (defeat learners) and 5 (marketsim)
- [Report 6](./manual_strategy/manual_strategy.md)
- No report for project 7
- [Report 8](./strategy_evaluation/strategy_evaluation.md)