pratmo / dbscan-airline-customers

Using DBSCAN for clustering and identifying the outliers to predict which East-West airline's frequent flier programme a consumer belongs to in order to target different types of mileage offers.

Contents:

• Coding Algorithm Series

  • Supervised Machine Learning Algorithms:
    • Regression - Linear Regression (using statsmodels and scikit-learn, ft. correlation, p-value analysis, multicollinearity, VIF, residual and regressor plots, model deletion diagnostics like cook's distance and H.influence points)
    • Classification -
  • Unsupervised Machine Learning Algorithms:
    • Dimensionality Reduction -
    • Clustering -
    • Association Rules -

• Projects

• Mini-Projects

Connect with me: