MDOD: Multi-Dimensional Data Outlier Detection

Introduction

A Python library for multi-dimensional data outlier and anomaly detection using advanced machine learning algorithms.

Publications

• "Outlier Detection Using Vector Cosine Similarity by Adding a Dimension" – Published in ICAIIC 2024 (DOI: 10.1109/ICAIIC60209.2024.10463442)
• I am looking for expert's endorsement to submit an article to the cs.AI section of arXiv. If you would like to endorse me, please visit the URL: (https://arxiv.org/auth/endorse?x=9QAAWI)

Installation

• Via pip: pip install mdod
  
  Or via GitHub:
  

  git clone https://github.com/mddod/mdod.git
  cd mdod
  python setup.py install

Usage Example

• import numpy as np
  import mdod
  
  localFile = 'TestDataset.txt'
  dets = np.loadtxt(localFile, delimiter=',')
  nd = 1  # Value of the observation point in the new dimension
  sn = 15  # Number of statistics on top scores
  result = mdod.md(dets, nd, sn)
  print(result)

  Download testmdod.py

Input Data Format (TestDataset.txt)

• data1,data2,data3,data4,data5,data6
  data1,data2,data3,data4,data5,data6
  data1,data2,data3,data4,data5,data6
  ...

  Download TestDataset.txt

Output Result Format

• [value1, '[data1 data2 data3 data4 data5 data6]', '0']
  [value2, '[data1 data2 data3 data4 data5 data6]', '1']
  [value3, '[data1 data2 data3 data4 data5 data6]', '2']
  ...

Additional Examples

• Random 2D Data: Download testmdod2D.py
  
• Random 3D Data: Download testmdod3D.py
  

Resources

For more information, visit GitHub or PyPI.