Skip to content

shivareddy2002/car-price-prediction

Repository files navigation

AutoPredict Pro is an end-to-end Machine Learning & Deep Learning web application that predicts the market value of used cars based on technical specifications, usage details, and seller information.
The system delivers accurate, data-driven price estimates through a professional Streamlit interface.


🔗 Live Demo:
👉 https://car-price-prediction-pro.streamlit.app/


📌 Problem Statement

Pricing used cars manually is challenging due to multiple influencing factors such as:

  • 📅 Manufacturing year & depreciation
  • 🛣️ Mileage and vehicle usage
  • ⚙️ Engine capacity & power output
  • ⛽ Fuel type & transmission mode
  • 👤 Ownership history & brand value

These complexities often lead to inconsistent pricing and poor decision-making.
AutoPredict Pro solves this using AI-driven models to provide fair and reliable car price predictions.


🏗️ System Architecture

The application follows a structured pipeline to ensure data integrity and prediction accuracy:

  1. 🧾 User Input – Features collected via Streamlit UI
  2. Data Validation – Logical consistency checks
  3. 🔄 Preprocessing – One-Hot Encoding & Feature Scaling (saved pipeline)
  4. 🧠 Model Inference – Random Forest or ANN prediction
  5. 💰 Output – Price inverse-transformed and displayed in ₹ INR

🧠 Models Used

1️⃣ Random Forest Regressor (Machine Learning 🌲)

  • Handles non-linear relationships effectively
  • Robust to outliers and reduces overfitting
  • Best suited for: General market trends & mid-range vehicles

2️⃣ Artificial Neural Network (Deep Learning 🤖)

  • Multi-layer dense neural network (TensorFlow/Keras)
  • Learns complex hidden patterns in high-dimensional data
  • Best suited for: Premium and modern vehicles

🛠️ Tech Stack

🎨 Frontend & Deployment

  • Streamlit – Interactive web dashboard
  • 🎨 HTML & CSS – Custom glass-morphism styling
  • 🎞️ Lottie Animations – Smooth UI animations

⚙️ Backend & Machine Learning

  • 🐍 Python – Core programming language
  • 🌳 Scikit-learn – Random Forest & preprocessing pipeline
  • 🧠 TensorFlow / Keras – Deep Learning ANN
  • 📊 Pandas & NumPy – Data manipulation & numerical computation

🔍 Features

✅ Core Features

  • 🔁 Dual-Model Prediction – Compare ML vs DL results in real time
  • 🇮🇳 INR Formatted Output – Prices displayed in ₹ for Indian market
  • 🎨 Professional UI – Clean, modern & responsive Streamlit design

📊 Market Analysis & Explainable AI (XAI)

  • 📌 Feature Importance – Impact of Year, Power, KM Driven, etc.
  • 📉 Depreciation Trends – How vehicle value decreases over time
  • 📈 Market Demand Insights – Fuel & transmission popularity analysis

🧪 Input Parameters

🔢 Feature 📝 Description
🚘 Brand Car manufacturer (Maruti, BMW, Hyundai, etc.)
📅 Year Manufacturing year
⛽ Fuel Type Petrol / Diesel / CNG / LPG
🔧 Transmission Manual / Automatic
🛣️ KM Driven Total distance covered
⚙️ Engine Engine capacity (CC)
⚡ Max Power Power output (BHP)
👤 Owner Number of previous owners
🏪 Seller Type Individual or Dealer

🔮 Future Enhancements

  • 📍 Location Intelligence – City/RTO-based price adjustments
  • 📸 Computer Vision – Damage detection from car images
  • 📡 Live Market Data – Real-time price updates via APIs
  • 🏢 B2B Dashboard – Bulk valuation tools for dealerships

📌 Project Workflow

flowchart LR
    A[Importing Libraries] --> B[Loading Car Price Dataset]
    B --> C[Data Cleaning & Feature Engineering]
    C --> D[Preprocessing: Encoding & Scaling]
    D --> E[Model Training: Random Forest / ANN]
    E --> F[Model Evaluation & Selection]
    F --> G[Prediction Output]
    G --> H[Deployment: Streamlit Web App]

    %% Styles
    style A fill:#FFD54F,stroke:#F57F17,stroke-width:2px,color:#000;
    style B fill:#4FC3F7,stroke:#0277BD,stroke-width:2px,color:#fff;
    style C fill:#AED581,stroke:#33691E,stroke-width:2px,color:#000;
    style D fill:#FFCC80,stroke:#EF6C00,stroke-width:2px,color:#000;
    style E fill:#BA68C8,stroke:#4A148C,stroke-width:2px,color:#fff;
    style F fill:#81D4FA,stroke:#01579B,stroke-width:2px,color:#000;
    style G fill:#FF8A65,stroke:#BF360C,stroke-width:2px,color:#fff;
    style H fill:#90CAF9,stroke:#0D47A1,stroke-width:2px,color:#000;
Loading

👨‍💻 Author

Lomada Siva Gangi Reddy

  • 🎓 B.Tech CSE (Data Science), RGMCET (2021–2025)
  • 💡 Interests: Python | Machine Learning | Deep Learning | Data Science
  • 📍 Open to Internships & Job Offers

Contact Me:


About

AutoPredict Pro – An end-to-end ML & Deep Learning web app for accurate used car price prediction using Random Forest & ANN, deployed with Streamlit.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors