Resumo
Machine Learning (ML) has been advancing in the most diverse areas of knowledge. Among those, ML for Healthcare and Decision Support of Medical applications present specific challenges in evaluating, monitoring, and maintaining ML models. Once deployed, models are subject to performance degradation (drift). Therefore, continuous monitoring and evaluation are essential to establish minimum performance guarantees over unknown, real-world data. In medical applications, incorrect decisions can lead to life-threatening situations and irreversible outcomes. The present work proposes a process for ML model evaluation designed for Healthcare applications running on real-world data. To that end, a conducted Systematic Literature Review (SLR) aimed at determining the state-of-the-art techniques and methods for ML evaluation (detailed in another paper) and a case study applied the proposed process to ML models in an oncology ICU. The Case Study produced positive outcomes in establishing a feedback loop for models in use against real-world data.
DOI:https://doi.org/10.56238/sevened2024.001-049