reg-visualization-tools

Visualization Tools for Comparative Analysis of Regression Models

Installation

pip install -r requirements.txt

To generate the plots as shown in the ICTAI article, please use the following command:

python ictai.py -ap

The functions defined use a DataFrame containing the predictions and the errors of the models, in the following format:

target	target_model1	error_model1	target_model2	error_model2
target1	value1	value2	...	...
target2	value1	value2	...	...
...	...	...	...	...

For instance, for two models predicting the price of houses:

price	price_model1	error_model1	price_model2	error_model2
154000	155400	1400	132600	-21400
98450	98420	-30	109500	11050

The script train.py can be used to train them:

python train.py --data <dataset> --target <target>