Abstract
Analysis of the evolution of the menstrual cycle symptomatology through Arima models. Despite the number of studies devoted to investigating the symptomatology of the menstrual cycle, little is known about how this symptomatology behaves. Within this context, this paper attempts to shed some light into the menstrual cycle symptomatology by analyzing women's symptomatologic sequences. Data were gathered by the daily recording of the menstrual symptoms. Time Series Analysis models (ARIMA models) were applied to analyze the data. Results showed that menstrual symptoms could be adjusted by means of simple autoregressive models in a 43.36% of the women, and by means of simple and cyclic autoregressive models in 26.3% of the women. For the rest of the women both mixed autoregressive and moving average models were applicable. The results reported here suggest that ARIMA models should be considered powerful analysis tools to describe and/or predict certain of human behaviors; and furthermore, menstrual symptomatology is defined as autoregressive (simple and/or cyclic).
