The Brownian Integral Kernel:

A New Kernel for Modeling Integrated Brownian Motions

This is the repo containing code, data, and supplementary material for the accompanying publication.

Installation:

install pyenv dependencies:

sudo apt update
sudo apt install -y make build-essential libssl-dev zlib1g-dev libbz2-dev \
libreadline-dev libsqlite3-dev wget curl llvm libncurses5-dev libncursesw5-dev \
xz-utils tk-dev libffi-dev liblzma-dev python-openssl git

install pyenv:

curl -L https://github.com/pyenv/pyenv-installer/raw/master/bin/pyenv-installer | bash

add it to bash, by adding the following to the bottom of the file ~/.bashrc:

export PATH="/home/user/.pyenv/bin:$PATH"
eval "$(pyenv init -)"
eval "$(pyenv virtualenv-init -)"

install python version:

pyenv install 3.9.6

install poetry:

curl -sSL https://install.python-poetry.org | python3 -

add it to bash, by adding the following to the bottom of the file ~/.bashrc:

export PATH="/home/[user]/.local/bin:$PATH"

set poetry to use pyenv:

poetry config virtualenvs.prefer-active-python true

And make sure venv are created inside a project:

poetry config virtualenvs.in-project true

install project dependencies:

poetry install

wait for all dependencies to install and you are finished.

You can now simply import and use the 'IntegralBrown' as you would with any other GPy kernel:

from brownian_integral_kernel.integral_kernel import IntegralBrown

... #Data loading

k = IntegralBrown(variance=1)
model = GPy.models.GPRegression(times, observation, k, noise_var=0.0)    

Experiments and Evaluation

After the installation you can easily reproduce all experiments, evaluation and figures:

1. Simply run the "exp_..." scripts which will run the experiments on the datasets or on generated data.
2. Run the "eval_..." scripts to calculate the evaluation metrics.
3. Run the "vis_..." scripts to generate the plots seen in the paper.

And you are finished.

We also provide all load profile data in the "LP" subfolder.