HousePrice Analytics

An Australian residential property-market dashboard exploring growth signals, modelled estimates, and composite investment scoring. Data freshness and source migration are under review.

🐍 Python📊 Streamlit🌲 XGBoost / LightGBM📈 Plotly🗄️ SQLite🐳 Docker🏛️ ABS 6416.0

Live dashboard · data review

🔗 houseprices.wfdnelson.com ↗

Capital Cities

160

Data Points

ML Models

Historical Data

What It Does

🔥

Growth Hotspot Detection

Statistical outlier analysis across 8 Australian capitals. Flags cities with YoY growth above market average + accelerating 3-quarter momentum. Scatter plots reveal acceleration vs year-on-year movement at a glance.

🤖

ML Price Prediction

XGBoost + LightGBM ensemble trained on 5 years of ABS quarterly index data. Features include 4-quarter lag values, rolling averages, and acceleration metrics. Time-series cross-validated for real predictive validity — not just in-sample fitting.

📊

Investment Scoring

Composite 0–100 score per city combining: growth momentum (40%), ML prediction confidence (30%), volatility-adjusted stability (15%), and market position signals (15%). Ranked table with radar chart breakdown for top cities.

📈

Trend Analysis

Interactive price index trends with city comparison, QoQ change bars, and YoY rankings. The current dashboard was built from historical ABS residential-property data. Its source pipeline is being reviewed against current ABS publications.

Stack

🏗️ Infrastructure

Docker (containerized app)
DigitalOcean droplet
Nginx reverse proxy
Refresh pipeline under review

🧠 ML Pipeline

scikit-learn (preprocessing, CV)
XGBoost + LightGBM ensemble
TimeSeriesSplit cross-validation
14-feature lag engineering

📡 Data

Historical ABS property data
Domain API (suburb stats)
SQLite database
Plotly interactive charts

Why I Built It

I built HousePrice Analytics as a practical way to move beyond ML tutorials and work through a complete modelling problem: choosing a useful question, preparing time-series data, engineering features, comparing models, and turning the results into an interface people can explore.

Property data made the trade-offs visible. Historical trends are easy to chart, but useful predictions require careful validation, honest treatment of uncertainty, and a clear distinction between a modelled signal and financial advice. The project became a hands-on learning environment for those decisions.

What I Learned

Time-series CV matters. Standard k-fold leaks future data into training — TimeSeriesSplit prevents this and gives a realistic performance estimate.
ABS data is surprisingly usable. The 6416.0 RPPI dataset is clean, consistent, and goes back decades. More reliable than scraped sources for trend analysis.
Ensemble prediction improves stability. XGBoost and LightGBM tend to diverge on edge cases — averaging their predictions reduces individual model bias.
Composite scoring needs weighting discipline. Easy to game a score by tweaking weights. Each weight needs a defensible reason — documented in the methodology.

← Back to Coding Projects 🔗 Live Dashboard ↗Similar: DAJobs Analytics →

HousePrice Analytics

An Australian residential property-market dashboard exploring growth signals, modelled estimates, and composite investment scoring. Data freshness and source migration are under review.

🐍 Python📊 Streamlit🌲 XGBoost / LightGBM📈 Plotly🗄️ SQLite🐳 Docker🏛️ ABS 6416.0

What It Does

🔥

Growth Hotspot Detection

🤖

ML Price Prediction

📊

Investment Scoring

📈

Trend Analysis

Stack

🏗️ Infrastructure

Docker (containerized app)
DigitalOcean droplet
Nginx reverse proxy
Refresh pipeline under review

🧠 ML Pipeline

scikit-learn (preprocessing, CV)
XGBoost + LightGBM ensemble
TimeSeriesSplit cross-validation
14-feature lag engineering

📡 Data

Historical ABS property data
Domain API (suburb stats)
SQLite database
Plotly interactive charts

Why I Built It

What I Learned

Time-series CV matters. Standard k-fold leaks future data into training — TimeSeriesSplit prevents this and gives a realistic performance estimate.

ABS data is surprisingly usable. The 6416.0 RPPI dataset is clean, consistent, and goes back decades. More reliable than scraped sources for trend analysis.

Ensemble prediction improves stability. XGBoost and LightGBM tend to diverge on edge cases — averaging their predictions reduces individual model bias.

Composite scoring needs weighting discipline. Easy to game a score by tweaking weights. Each weight needs a defensible reason — documented in the methodology.