Phone Price Segmentation using Machine Learning

• Train Dataset was categorised in 4 groups of price range

  • 0 (low cost)
  • 1 (medium cost)
  • 2 (high cost)
  • 3 (very high cost)

• Model was trained on 20 different features such as:

  • battery power
  • dual sim (has or not)
  • front camera megapixels
  • ram
  • wifi
  • 4G or not etc.

• Data Preprocessing

  • Pandas
  • Numpy
  • Standard Scaler

• Data Visualisation

  • Correlation heatmap
  • Seaborn
  • Matplotlib

• Classifiers

  • Random Forest
  • Support Vector Machine

Random Forest outperforms on train dataset. A comparison of both model prediciton shown in plot.