pkmap

a dataset visualization generator supposed to exhibit the imbalance for human eye acquisition


Keywords
disaggregation, imbalanced-data, karnaugh-map, metrics, nilm, observability, pseudo-karnaugh-map, visualization
License
MPL-2.0
Install
pip install pkmap==0.2.1

Documentation

build PyPI python license

the Pseudo Karnaugh Map

Introduction

The Pseudo Karnaugh Map (PKMap), a coloured Weighted / Multi-valued Karnaugh Map, an Enchanced Karnaugh Map as the project's initially named, is an evaluation method to visualize the imbalanced class distribution of a dataset.

This project was started to guide the generating of synthesis datasets in which a self-collected data is added into a public dataset. Later, we found it works in dataset imbalance representation as well.

Why PKMap

Gain instructive insight of a datset before utilizing is currently virtually impossible. The PKMap is expected to offer an intuitionistic and visualized recognization of the imbalanced classes distribution of a dataset.

Documantation

PKMap Documantation

Publications

A primiary paper about the PKMap has been publish, more is on the way.

  1. Lu Z, Liu G, Liao R. A pseudo Karnaugh mapping approach for datasets imbalance. In: E3S Web of Conferences. Vol 236. EDP Sciences; 2021:04006. DOI:10.1051/e3sconf/202123604006

Here is the BibLaTeX code:

@inproceedings{lu2021,
  title = {A Pseudo {{Karnaugh}} Mapping Approach for Datasets Imbalance},
  booktitle = {{{E3S Web}} of {{Conferences}}},
  author = {Lu, Zhijian and Liu, Gang and Liao, Rongwen},
  editor = {Anpo, M. and Song, F.},
  date = {2021},
  volume = {236},
  pages = {04006},
  publisher = {{EDP Sciences}},
  doi = {10.1051/e3sconf/202123604006},
  eventtitle = {3rd {{International Conference}} on {{Energy Resources}} and {{Sustainable Development}} ({{ICERSD2020}})}
}