A Supervised Learning Regression Method for the Analysis of the Taste Functions of Healthy Controls and Patients with Chemosensory Loss

Research output: Contribution to journalResearch articleContributedpeer-review



In healthy humans, taste sensitivity varies widely, influencing food selection and nutritional status. Chemosensory loss has been associated with numerous pathological disorders and pharmacological interventions. Reliable psychophysical methods are crucial for analyzing the taste function during routine clinical assessment. However, in the daily clinical routine, they are often considered too time-consuming. We used a supervised learning (SL) regression method to analyze with high precision the overall taste statuses of healthy controls (HCs) and patients with chemosensory loss, and to characterize the combination of responses that would best predict the overall taste statuses of the subjects in the two groups. The random forest regressor model allowed us to achieve our objective. The analysis of the order of importance of each parameter and their impact on the prediction of the overall taste statuses of the subjects in the two groups showed that salty (low-concentration) and sour (high-concentration) stimuli specifically characterized healthy subjects, while bitter (high-concentration) and astringent (high-concentration) stimuli identified patients with chemosensory loss. Although the present results require confirmation in studies with larger samples, the identification of such distinctions should be of interest to the health system because they may justify the use of specific stimuli during the routine clinical assessments of taste function and thereby reduce time and cost commitments.


Original languageEnglish
Article number2133
Issue number8
Publication statusPublished - Aug 2023

External IDs

ORCID /0000-0001-9713-0183/work/146645738



  • general taste status, random forest regressor, supervised learning regression, taste loss