This project aims to predict real estate prices per square meter in different regions of Sofia, Bulgaria. By leveraging historical pricing data and various features, we build predictive models that can help in understanding the trends and making informed decisions in the real estate market.
The main goal of this project is to develop a machine learning model that accurately predicts the price per square meter of apartments in Sofia based on different factors such as location, time, and other relevant attributes. The project is organized into several steps, including data collection, exploration, feature engineering, model training, and evaluation.
Predicting Sofia Real Estate Prices.ipynb: This is the main notebook where the model is trained and evaluated. It provides a comprehensive walkthrough of the data preprocessing, feature engineering, and model selection processes. Start with this file to understand the overall flow and methodology used in the project.
-
first_dataset_eda_and_modeling.ipynb: Contains the exploratory data analysis (EDA) and initial modeling based on the first dataset. It explores different features and their impact on the model.
-
second_dataset_eda_and_modeling.ipynb: Similar to the first notebook, this file handles EDA and modeling for the second dataset. It includes additional testing of models and feature refinements.
-
data_scraping.ipynb: This notebook is dedicated to the data collection process, where the raw data is scraped from relevant sources and prepared for analysis.
-
data/: Contains the original datasets used in the project. The data has been collected and preprocessed to ensure consistency and accuracy.
-
data2/: Includes additional data that was added later to enhance the model's performance and increase the accuracy of predictions.
-
Start by exploring the Predicting Sofia Real Estate Prices.ipynb notebook. This notebook is your go-to guide for understanding the overall project structure, methodologies used, and the final model outcomes.
-
Dive into the EDA notebooks (first_dataset_eda_and_modeling.ipynb and second_dataset_eda_and_modeling.ipynb) to see the detailed exploratory analysis and the rationale behind feature selection and model refinement.
-
Check out data_scraping.ipynb if you're interested in how the data was gathered and prepared.
- Data Updates: New data added to improve model accuracy.
- Model Training: Ongoing improvements and testing of different models to enhance predictive performance.
- Further Model Optimization: Continue refining the model with additional data and advanced techniques.
- Deployment: Explore deployment options for the predictive model, potentially creating a user-friendly interface for real estate stakeholders in Sofia.