Co-Regularised Least Squares Regression

This Repo Contains python implementation of paper Efficient Co-Regularised Least Squares Regression - by Ulf Brefeld, Thomas G¨artner, Tobias Scheffer and Stefan Wrobel

Instructions

Run the below line

python train.py --dataset <path_to_dataset> --epochs <no of epochs>

--outputDim is use to specify the number of output dimension

• The model will take last outputDim column of dataset as target labels

Baseline

For comparison, we are implementing RSE as baseline (also known as Ridge regression)

Example

Step1: keep the processed dataset file at ./dataset with last column corresponds to target label

step2:

$ python train.py --dataset ./dataset/pollution.data.txt --epochs 9000 --batch_size 16 --outputDim 1

split of attribute at
9
Loading the datasets...
model...
File name pollution.data.txt
Starting training for 9000 epoch(s)
Training completed for printing loss uncomment 75 and 76 linn in train.
Starting testing for epoch(s)
combine-rms on test set: 51.64
done
Training of Baseline starting
Training of Baseline successful. To print losses uncomment the line 67 and 68
Baseline testing Starting
baseline loss on test: 89.166

Miscellanous

• Lisence: MIT
• You may contact me by opening an issue on this repo. Please allow 2-3 days of time to address the issue.