Project information

Summary: This project focuses on an unsupervised learning technique called self - organizing maps, utilized to visualize the 13 dimensional wine data of 3 classes and map onto a 2D grid/map. It uses RGB colors to visually separate the wine data based on its class. Then we use this map as blueprint to classify a test dataset and achieved 94% accuracy.

Legend:

  • Red: Class 1
  • Green: Class 2
  • Blue: Class 3
  • Black: No Entry