Lessons from a human-in-the-loop machine learning approach for identifying vacant, abandoned, and deteriorated properties in Savannah, Georgia

TL;DR

This paper presents VADecide, a human-in-the-loop machine learning model for identifying vacant, abandoned, and deteriorated properties, demonstrating improved accuracy and insights into human-machine differences in urban planning applications.

Contribution

It introduces VADecide, a novel HITLML model for VAD property identification, showing enhanced accuracy and revealing differences between human and machine predictions.

Findings

Higher prediction accuracy with HITLML compared to non-human-involved models

Revealed differences between human and machine-generated results

Contributed insights into HITLML advantages and challenges in urban planning

Abstract

Addressing strategies for managing vacant, abandoned, and deteriorated (VAD) properties is important for maintaining healthy communities. Yet, the process of identifying these properties can be difficult. Here, we create a human-in-the-loop machine learning (HITLML) model called VADecide and apply it to a parcel-level case study in Savannah, Georgia. The results show a higher prediction accuracy than was achieved when using a machine learning model without human input in the training. The HITLML approach also reveals differences between machine vs. human-generated results. Our findings contribute to knowledge about the advantages and challenges of HITLML in urban planning. [Accepted for Publication at a Peer Review Journal]