Parul Agarwal
Slum Prediction Model

Using arcGIS and satellite images, I studied the spatial distribution of urban features in existing slums in India. This information helped me create a prediction model that could predict where urban slums may develop in the future. A predictive model is necessary because slums are formed through unplanned urban expansion.

Hypothesized causal pathway for RCT

Predicting future slum locations and growth will allow policymakers, government officials, and NGOs to create a plan, whether it is through affordable housing, healthcare, education, or basic infrastructure for urban growth.

The project was done into two parts, a macro exploration that looked at India as a whole and a micro exploration that looked at a subsection of Mumbai. For Mumbai's subsection, my predictive variables were railways, highways, greenspace, and railroad cross (four-point) or star (five-point) intersections.