Fig. 1

Study overview. The participants were examined twice, once during the follicular phase (within five to ten days after the onset of menstrual bleeding) and once during the late-luteal phase (within five days before the next menstruation). For each participant, the data from each phase were counted as one data point. In total, 121 data points were obtained from 60 participants. General questions, the Premenstrual Symptom Screening Tool, Sense of Coherence (SOC), and the Parental Bonding Instrument (PBI) were completed once; the other parts of the questionnaires were completed at both the follicular and late-luteal phases. All responses were converted to numerical or categorical variables, and free descriptions were removed. Spectral analysis of heart-rate variability (HRV) was performed for the follicular and late-luteal phases. To obtain the binary objective variable, the Japanese version of the Premenstrual Symptom Screening Tool (PSST) was applied, placing participants with moderate-to-severe premenstrual syndrome (PMS) and premenstrual dysphoric disorder (PMDD) in the “PMD” group and those with other conditions in the “non-PMD” group. The explanatory variables comprised 443 features. PMD, Premenstrual disorder