Data Analysis of Missing People in Ecuador

Abstract

The phenomenon of missing people has negative consequences for the individuals affected, their families, and society at large. This issue has become increasingly prominent in recent years, partly due to the influence of social media, which has emerged as a vital tool in the search for missing persons. Given the growing importance of addressing this challenge, this research contributes to the expanding field of machine learning applications for social issues. Supervised machine learning models were applied to open data on missing persons in Ecuador between 2021 and 2024 to predict the status of individuals as either "Found" or "Deceased." Using personal, social, and event-specific variables, two models Random Forest (RF) and Support Vector Machine (SVM) were implemented and evaluated. The models were assessed using key performance metrics, including accuracy, precision, recall, F1 score, and confusion matrices, to determine their effectiveness. The analysis revealed that the RF model achieved superior performance on the test data compared to SVM, with an accuracy of 89%, demonstrating its suitability for the dataset. These findings provide valuable insights into the factors influencing the outcomes of disappearance cases, allowing decision-makers to optimize resource allocation, improve search strategies, and support evidence-based decision-making. Predicting the status of a missing person offers an innovative approach to addressing this critical social issue.