Airbnb Toronto Price Prediction

The AI-driven price prediction model aims to predict the prices of Airbnb listings in Toronto using three models: Ridge Regression, Light GBM, and Multi-layered Perceptron (Feed-forward Neural Network). The application is built using Streamlit and various data science libraries.

Table of Contents

• Process
• Models Evaluation
• Installation
• Usage
• Project Structure
• Contributing
• License

Process

Data Ingestion

Dataset containing listings for Toronto available via Inside Airbnb. The dataset contains 20k+ listings with 74 different characteristics.

Initial Feature Elimination

Removed any unnecessary columns, renamed columns for better readability and converted price column to float.

Exploratory Data Analysis (EDA)

Explored various features of data to identify patterns and visualized using different graphs.

Data Preprocessing

Train-test split the data by 90-10 percentages

Took natural log of price

Handled missing values through imputation and removal depending on requirements.

Engineered new features such as days_since columns for host registration

Engineered new features for distance to downtown

Encoded categorical columns using boolean, ordinal and one-hot encoding depending on requirements

Modeling and Evaluation

Built 3 different models.

Ridge Regression with regularization

LightGBM with correct parameters to prevent overfitting

Multi-layered Perceptron (MLP) with 3 hidden layers

Evaluated using MSE and R2 scores

Model Interpretation

Used SHAP for local and global interpretations

Models Evaluation

Installation

1. Clone the repository:

   git clone /~https://github.com/rajprasadshrestha/AirbnbDyanmicPricingOfToronto.git
   cd AirbnbDyanmicPricingOfToronto

2. Create a virtual environment:

   python3 -m venv venv
   source venv/bin/activate

3. Install the required packages:

   pip install -r requirements.txt

Usage

1. Run the Streamlit app:

   streamlit run app.py

2. Access the application:

   Open your web browser and go to http://localhost:8501.

3. Online Access:

   You can also access the application online at this link.

Project Structure

• app.py: Main application file.
• requirements.txt: List of required Python packages.
• dataset/: Directory containing Airbnb listing dataset.
• backend/: Jupyter notebooks for data exploration and model training

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any changes.

License

This project is licensed under the MIT License. See the LICENSE file for details.