The pivotal role of interpretability in employee attrition prediction and decision-making

This article explores the evolution of machine learning (ML) algorithms, emphasizing the growing importance of interpretability in understanding automated decisions. Progress from early to advanced ML models highlights the need for better performance and adaptability. However, the inherent black-box...

Descripción completa

Detalles Bibliográficos
Autores: Marín Díaz, Gabriel, Galán Hernández, José Javier
Tipo de recurso: capítulo de libro
Fecha de publicación:2024
País:España
Institución:Universidad Complutense de Madrid (UCM)
Repositorio:Docta Complutense
Idioma:inglés
OAI Identifier:oai:docta.ucm.es:20.500.14352/130072
Acceso en línea:https://hdl.handle.net/20.500.14352/130072
Access Level:acceso abierto
Palabra clave:004.85
519.816
519.22-7
658
658.3
Decision-making
Machine learning
XAI
Interpretability
AI
AHP
Inteligencia artificial (Informática)
Teoría de la decisión
Estadística matemática (Matemáticas)
Investigación operativa (Estadística)
Administración de empresas
1203.04 Inteligencia Artificial
1209.04 Teoría y Proceso de decisión
5311.04 Organización de Recursos Humanos
5311.07 Investigación Operativa
Descripción
Sumario:This article explores the evolution of machine learning (ML) algorithms, emphasizing the growing importance of interpretability in understanding automated decisions. Progress from early to advanced ML models highlights the need for better performance and adaptability. However, the inherent black-box nature of many ML algorithms raises challenges, underscoring the necessity for interpretability to improve transparency and accountability. Examining the evolution of interpretability in ML, the article showcases advancements in techniques facilitating human comprehension of decision-making processes. As ML becomes integral across domains, the article underscores the importance of interpretable models to bridge the gap between automated decisions and human understanding. The article delves into the changing role of humans in decision-making. Despite the efficiency of ML algorithms, the interpretability factor prompts a revaluation of human involvement, necessitating a balanced approach for ethical AI deployment. Furthermore, the article explores integrating decision-making methods like Analytic Hierarchy Process (AHP) to enhance interpretability. Proposing a framework that combines AHP with interpretable ML models, it suggests a structured approach for human-in-the-loop decision-making while considering feature importance.