lightonai / newma

Implementation of NEWMA: a new method for scalable model-free online change-point detection

NEWMA

A new method for scalable model-free online change-point detection.

This repository contains the code for NEWMA: a new method for scalable model-free online change-point detection, Nicolas Keriven, Damien Garreau, Iacopo Poli.

To cite this work

@ARTICLE{9078835, 
author={N. {Keriven} and D. {Garreau} and I. {Poli}}, 
journal={IEEE Transactions on Signal Processing}, 
title={NEWMA: a new method for scalable model-free online change-point detection}, 
year={2020}, 
volume={}, 
number={}, 
pages={1-1},} 

Code

Requirements

The code is written for Python 3.
You can install the Python modules required by running pip install -r requirements.txt inside the folder.

Installing the onlinecp package

You can install the onlinecp Python package by running pip install ./ from the root folder of this repository.

Access to Optical Processing Units

To request access to LightOn Cloud and try our photonic co-processor, please visit: https://cloud.lighton.ai/

For researchers, we also have a LightOn Cloud for Research program, please visit https://cloud.lighton.ai/lighton-research/ for more information.

Figures in the paper

You can generate data for the figures in the paper as follows:

• run test_dim.py and test_B_runningtime.py for Figure 4a
• run test_adaptive_vs_fixed.py for Figure 4b
• run test_algos_synthetic_data.sh for Figure 4c
• run test_algos_vad.sh for Figure 4d

The scripts to generate the plots from data are in plots and they have the same name prepended by plot_. Look at plots/README.md for info on how to run them.

Code for old version of the paper (v1)

The code for the older version of our paper is in code_v1. The subdirectory contains its README.md.