July 2023 in “Dermatology practical & conceptual” This study developed a support vector machine model using trichoscopic patterns to accurately classify androgenic alopecia severity, with an accuracy of 94.3% in training and 90.0% in test datasets.
5 citations
,
March 2022 in “Clinical Cosmetic and Investigational Dermatology” This study proposed a model that accurately predicts skin condition using genotype information and machine learning, suggesting potential for creating customized cosmetics.
July 2025 in “Journal of Investigative Dermatology” This study found that both desmoglein-specific and non-desmoglein autoantibodies may play active roles in Pemphigus vulgaris pathogenesis, with HLA genetics influencing autoimmune specificity.
20 citations
,
September 2020 in “International journal of computer applications” This study found that the Random Forest machine learning algorithm achieved the highest accuracy, 96%, in diagnosing Polycystic Ovarian Syndrome using patients' clinical data.
2 citations
,
November 2024 This review discussed recent research on using machine learning to predict mental disorders, reporting that Adaboost could predict depression with 92.5% accuracy and 93.6% specificity, while other models like XGBoost and RNN were applied for post-stroke depression and EEG-based depression detection, respectively.