Mobile Phone Price Prediction Project

This project focuses on predicting the price range category of mobile phones based on various hardware and connectivity features. Using a dataset containing specifications such as battery power, RAM, camera megapixels, screen size, connectivity options and many more, we train multiple machine learning models to classify phones into four price categories:

• 0 = low cost
• 1 = medium cost
• 2 = high cost
• 3 = very high cost

---

Key Highlights

• Dataset Features:
  The dataset includes the following mobile phone specifications:

  battery_power: Battery Capacity in mAh
  blue: Has Bluetooth or not
  clock_speed: Processor speed (GHz)
  dual_sim: Has dual SIM support or not
  fc: Front camera megapixels
  four_g: Has 4G or not
  int_memory: Internal Memory in GB
  m_dep: Mobile depth in cm
  mobile_wt: Weight in grams
  n_cores: Processor Core Count
  pc: Primary Camera megapixels
  px_height: Pixel Resolution height
  px_width: Pixel Resolution width
  ram: RAM in MB
  sc_h: Mobile Screen height in cm
  sc_w: Mobile Screen width in cm
  talk_time: Time a single battery charge will last, in hours
  three_g: Has 3G or not
  touch_screen: Has touch screen or not
  wifi: Has WiFi or not

• Feature Engineering:
  Created additional features such as:

  • resolution
  • total camera megapixels
  • battery-to-RAM ratio
  • memory-to-RAM ratio
  • core speed
  • connectivity count
  • talk time per battery power

• Models Evaluated:
  Logistic Regression, Stochastic Gradient Descent Classifier (SGD), Linear Discriminant Analysis (LDA), Random Forest, XGBoost, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Gaussian Naive Bayes.

• Final Model:
  Selected Logistic Regression for final deployment due to its high accuracy and interpretability, achieving over 97% accuracy on the test dataset.

• Prediction Output:
  The model outputs a price category label (e.g., low cost, medium cost, high cost, very high cost) for easy understanding.

---

Usage

• Provide mobile specifications as input features in the required format.
• The trained model predicts the price range category.
• The output prediction is displayed with a descriptive label for better user interpretation.

Python Version

• Python 3.12.5