Regression

This repository contains Python code to perform 1-D regression with:

• Gaussian Process Regression,
• Relevance Vector Machine.

Requirements

• Install the latest version of Python 3.X.
• Install the required packages:

pip install -r requirements.txt
pip install https://github.com/JamesRitchie/scikit-rvm/archive/master.zip

Usage

python main.py

Results

The ground truth is the sinc function.

Influence of the number of training samples

The variable noise_level is set to 0.1. The variable training_data_range is set to a large value (15).

The results are shown with increasing number of training samples.

Influence of the noise level

The variable num_samples is set to 100. The variable training_data_range is set to a large value (15).

The results are shown with increasing noise level.

Influence of range of training data

The variable num_samples is set to 100. The variable noise_level is set to 0.1.

The results are shown with increasing range of training data

References

• Python module scikit-learn
• Documentation: Gaussian Process with scikit-learn
• Python module scikit-rvm
• Python module sklearn-rvm
• Slides about Relevance Vector Regression
• Slides about Gaussian Process Regression