This project is a machine learning–driven REIT recommendation system that helps investors discover real estate investment trusts (REITs aligned with their preferences and financial goals.
The system models investor inputs such as risk tolerance, budget range, investment horizon, liquidity needs, preferred property type, and location, then matches them against an existing pool of REITs using constraint-based filtering and similarity scoring.
The focus of this project is clarity, explainability, and ML fundamentals, not production deployment.
Investors often struggle to identify REITs that match their:
- Risk appetite
- Capital constraints
- Time horizon
- Liquidity expectations
- Real estate exposure preferences
Most platforms either overwhelm users with data or provide opaque recommendations.
This project demonstrates how simple machine learning techniques can be used to generate transparent, preference-aware investment recommendations.
The system works in four main stages:
The user provides:
- Risk tolerance (low → high)
- Minimum and maximum investment amount
- Investment horizon (short / medium / long term)
- Liquidity preference
- Preferred property type (e.g. residential, commercial)
- Preferred location
These inputs represent the investor profile.
Each REIT in the dataset is described using structured features such as:
- Risk level
- Minimum investment
- Expected liquidity
- Property focus
- Geographic exposure
- Expected return profile
These become the features used for matching.
Before any ML scoring:
- REITs that fall outside the investor’s budget range are removed
- REITs that violate non-negotiable preferences are excluded
This ensures recommendations are realistic and relevant.
Using numerical feature vectors:
- Investor preferences are compared to each REIT
- Similarity scores are computed
- REITs are ranked from best match to weakest match
The top-ranked REITs are returned as recommendations.
- Preference-based investment recommendations
- Explainable similarity scoring
- Beginner-friendly ML logic
- Clear separation of filtering vs scoring
- Designed for easy extension
- Python
- Pandas
- NumPy
- Scikit-learn
The system returns:
- A ranked list of REIT recommendations
- A compatibility score for each REIT
- A plain-language explanation of why the top REITs were selected
This project is for educational purposes only and does not constitute financial advice.
- Add architecture diagrams
- Help you structure the Fintech Investment Advisor README next