Stanford CS221 Artificial Intelligence Project
Install dependencies in requirements.txt for Python 3.6+:
pip install -r requirements.txtUnzip data.zip and place kddcup.data_10_percent in resources/datasets/.
The dataset is taken from the UCI ML repository. Head over to that page and grab the kddcup.data_10_percent.gz file, uncompress, and place kddcup.data_10_percent in resources/datasets/:
mkdir -p resources/datasets
wget -O - http://kdd.ics.uci.edu/databases/kddcup99/kddcup.data_10_percent.gz | gunzip -c > resources/datasets/kddcup.data_10_percent
Run the experiments.py file as follows:
python experiments.pyThis option does not produce very visually appealing output. Alternatively you can take a look at final.ipynb with graphs and wider tables.
Note: experiments results are stored in the results folder. You can remove this folder to run everything from scratch. The experiments took less than 30 minutes to complete on a MacBook Pro 2.9 GHz Intel Core i7 with 16GB of memory.