Cost-sensitive learning for imbalanced medical data: a review

Abstract

Integrating Machine Learning (ML) in medicine has unlocked many opportunities to harness complex medical data, enhancing patient outcomes and advancing the field. However, the inherent imbalanced distribution of medical data poses a significant challenge, resulting in biased ML models that perform poorly on minority classes. Mitigating the impact of class imbalance has prompted researchers to explore various strategies, wherein Cost-Sensitive Learning (CSL) arises as a promising approach to improve the accuracy and reliability of ML models. This paper presents the first review of CSL for imbalanced medical data. A comprehensive exploration of the existing literature encompassed papers published from January 2010 to December 2022 and sourced from five major digital libraries. A total of 173 papers were selected, analysed, and classified based on key criteria, including publication years, channels and sources, research types, empirical types, medical sub-fields, medical tasks, CSL approaches, strengths and weaknesses of CSL, frequently used datasets and data types, evaluation metrics, and development tools. The results indicate a noteworthy publication rise, particularly since 2020, and a strong preference for CSL direct approaches. Data type analysis unveiled diverse modalities, with medical images prevailing. The underutilisation of cost-related metrics and the prevalence of Python as the primary programming tool are highlighted. The strengths and weaknesses analysis covered three aspects: CSL strategy, CSL approaches, and relevant works. This study serves as a valuable resource for researchers seeking to explore the current state of research, identify strengths and gaps in the existing literature and advance CSL’s application for imbalanced medical data.

Exploring Resampling Techniques in Credit Card Default Prediction

“…Imbalanced datasets present numerous challenges for machine learning algorithms (Araf et al, 2024). The skewed distribution of classes can introduce biases during model training, favouring the majority class and resulting in diminished performance for the minority class.…”

“…Imbalanced classi cation scenarios, also known as rare event modelling, arise when the target variable exhibits a signi cant imbalance, with the minority class representing the rare events. In such cases, the model's tendency to learn predominantly from the majority class can make predicting the minority class particularly di cult (Araf et al, 2024;Khushi et al, 2021;Naseriparsa & Kashani, 2014). Consequently, machine learning algorithms may struggle to construct accurate models (Chawla et al, 2002), leading to challenges in evaluation metrics that could fall into the "metric trap" and result in inaccurate results (Jeni et al, 2013;Yap et al, 2014).…”

“…Consequently, this can lead to consistently high-performance accuracy scores that are not necessarily reliable (An & Suh, 2020;Hooda et al, 2018;Lin et al, 2023). Relying solely on raw performance accuracy may not be ideal for evaluating classi cation models, especially in the context of imbalanced datasets (Araf et al, 2024;Hooda et al, 2018). This outcome is because accuracy only considers true positive/negative observations and ignores false positive and negative observations.…”