💻 Laptop Price Predictor

Ever wondered how much your dream laptop should cost? 🤔
This project is a Machine Learning model that predicts the price of a laptop based on its configuration — things like RAM, CPU, GPU, brand, OS, touchscreen, weight, and more.

I trained this model using a dataset of laptops and deployed it with a simple interface so you can test it yourself! 🚀

---

✅ Try it out here! 🚀

📊 Dataset & EDA

• Cleaned the dataset by dropping unnecessary columns like Cpu, Cpu Name, Memory (after feature engineering).
• Extracted useful features such as:
  • Whether the screen is Touchscreen or not
  • Whether it has an IPS panel
  • Screen resolution & PPI
  • Encoded categorical variables like Company, TypeName, OS, GPU Brand using OneHotEncoder.
• Handled missing values & ensured consistent formatting for numerical + categorical columns.
• Target variable: Laptop Price (log-transformed for stability).

---

⚙️ Model Building

• Used a Pipeline for preprocessing + model training.
• Steps:
  1. ColumnTransformer → One-hot encoding for categorical features.
  2. RandomForestRegressor as the main model.
• Hyperparameters tuned:
  • n_estimators=100
  • max_depth=15
  • max_features=0.75
  • max_samples=0.5

---

📈 Performance

• Evaluated on test data.
• 📌 R² Score: ~0.86
• 📌 Mean Absolute Error (MAE): ~0.19 (after taking exp back from log)

Pretty solid for predicting real-world laptop prices .

---

🛠️ Tech Stack

• Python
• Scikit-learn (ML model)
• Pandas, NumPy (data wrangling)
• Matplotlib, Seaborn (EDA & plots)
• Streamlit (for deployment / frontend)