This project predicts the price range of mobile phones based on features such as RAM, battery power, internal memory, camera quality, and screen dimensions.
It is a supervised machine learning classification problem designed to simulate real-world product pricing and customer segmentation.
Given mobile specifications, the objective is to classify devices into different price categories.
This helps in:
- Understanding market segmentation
- Assisting customers in choosing the best product
- Supporting pricing strategy decisions
- Source: Kaggle
- Training data: 2000 rows, 21 features
- Test data: 1000 rows, 21 features
Features include:
- RAM
- Battery power
- Camera specifications
- Screen size
- Processor cores
- Python
- Jupyter Notebook
- Pandas, NumPy
- Scikit-learn
- Matplotlib, Seaborn
- Data loading and validation
- Data cleaning and preprocessing
- Exploratory Data Analysis (EDA)
- Feature selection
- Train-test split (85% train, 15% test)
- Model training
- Model evaluation
- Model used: Random Forest Classifier
- The model performs well across most price categories
- High accuracy in predicting most price ranges
- Slight difficulty in identifying price range 2, but performs well for others
- Higher RAM → Higher price range
- Better battery → Higher price segment
- Positive correlation:
- Primary & front camera
- Screen width & height
- Customers prefer:
- 4G-enabled phones in higher price range
- More features as budget increases
This project predicts the price range of mobile phones based on features such as RAM, battery power, internal memory, camera quality, and screen dimensions.
It is a supervised machine learning classification problem designed to simulate real-world product pricing and customer segmentation.
Given mobile specifications, the objective is to classify devices into different price categories.
This helps in:
- Understanding market segmentation
- Assisting customers in choosing the best product
- Supporting pricing strategy decisions
- Source: Kaggle
- Training data: 2000 rows, 21 features
- Test data: 1000 rows, 21 features
Features include:
- RAM
- Battery power
- Camera specifications
- Screen size
- Processor cores
- Python
- Jupyter Notebook
- Pandas, NumPy
- Scikit-learn
- Matplotlib, Seaborn
- Data loading and validation
- Data cleaning and preprocessing
- Exploratory Data Analysis (EDA)
- Feature selection
- Train-test split (85% train, 15% test)
- Model training
- Model evaluation
- Model used: Random Forest Classifier
- The model performs well across most price categories
- High accuracy in predicting most price ranges
- Slight difficulty in identifying price range 2, but performs well for others
- Higher RAM → Higher price range
- Better battery → Higher price segment
- Positive correlation:
- Primary & front camera
- Screen width & height
- Customers prefer:
- 4G-enabled phones in higher price range
- More features as budget increases
Mobile-Price-Classification/
│ ├── mobile-price-classification-final.ipynb
├── train.csv
├── test.csv
├── README.md
- Clone the repository
- Install required libraries
- Open Jupyter Notebook
- Run all cells
- Add model comparison (Logistic Regression, SVM, etc.)
- Improve accuracy for mid-range category (price range 2)
- Convert notebook into modular ML pipeline
- Add deployment (Streamlit / Flask)
Pragya Kapil
QA Automation Engineer | Data Science & AI/ML