A data analytics project that forecasts residential construction values in the Philippines using various data analytics techniques. The purpose of this analysis is to evaluate and forecast the total value (in ₱1,000) of single housing under residential construction in the Philippines, with a focus on identifying key growth periods and high-performing regions from 2020 to 2024. The goal is to project the potential construction value for 2025, helping stakeholders gain foresight into upcoming trends in the residential sector.
Timeline: March 2025 - May 2025
Project Type: Academic Work
Skills Developed
- Research and Analytical Thinking
- Data Cleaning and Transformation
- Time Series Forecasting
- Data Visualization
- Interpretation of Regression Output
Link
Github
The WHAT
Objectives
- Forecast Total: Forecast the total value of residential construction in the Philippines for 2025 using historical data and analytics.
- Trend Analysis: Identify key growth periods and high-performing regions from 2020 to 2024 in the residential sector.
- Data-Driven Insights: Utilize various data analytics techniques to extract actionable insights for stakeholders.
- Strategic Foresight. Help stakeholders anticipate upcoming trends and make informed decisions in real estate development.
Key Metrics
- Total Value (₱1,000): Reflects the overall annual worth of single housing residential construction activity each year and serves as the foundation for identifying growth trends.
- Value per Region: Compares construction value across regions to pinpoint leaders in residential development and identify areas with rising or lagging activity.
- Year-on-Year Trend: Tracks changes in construction value each year to highlight growth patterns and shifts in the residential sector.
- Forecasted Value for 2025: Uses historical values from 2020–2024 and time-series forecasting to project the potential construction value for 2025.
The HOW
I collected and cleaned raw construction data, including permit counts and values, to perform regional time-series forecasting. I utilized various forecasting methods—Naïve, average of past values, 3-week moving average, and exponential smoothing—and selected the most accurate model based on MAE, MSE, and MAPE error metrics. The trends were then effectively visualized using Excel, incorporating line charts, bar graphs, and regression outputs.
Before
After
Tools
Key Findings
<aside>
💡
- Focus Investments in High-Growth Areas: Prioritize Region IV-A and surrounding regions for real estate projects, housing developments, and commercial infrastructure.
- Regional Infrastructure Support: Encourage public-private partnerships in low-activity regions like ARMM and Region XIII to stimulate construction and economic development.
- Policy Intervention: Recommend government support or incentives in underdeveloped regions to balance national growth and avoid over-concentration in already booming areas.
</aside>