/Fake_banknote_detection

Detecting fake money using machine learning

Primary LanguageJupyter Notebook

👉Notebook👈

💪 Learning goals:

  • apply and interpret PCA (Principal component analysis)
  • apply K-means clustering algorithm
  • build and apply logistic regression
  • build, train and apply Machine Learning algorithms

💼 Scenario:

We work for a govermental agency that fights counterfeiting. Our mission is to create a counterfeit detection algorithm.

📦 Dataset:

Our dataset contains geometric characteristics of banknotes. For each of them, we know:

  • length of the banknote (in mm) ;
  • height of the banknote (measured on the left side, in mm);
  • height of the banknote (measured on the right side, in mm) ;
  • margin between the top edge of the banknote and the banknote image (in mm);
  • margin between the lower edge of the banknote and its image (in mm);
  • diagonal of the banknote (in mm).

📊 Results and conclusions:

Full presentation

Relationships between variables

PCA projection

Description of each cluster