A comparison of popular fertility awareness methods to a DBN model of the woman's monthly cycle

Fertility Awareness Methods are effective, safe, and low-cost techniques for identifying the fertile days of a menstrual cycle. In this paper, we compare the effectiveness of predicting the fertile days by a Dynamic Bayesian Network model of the monthly cycle to 11 existing Fertility Awareness Methods. We base our comparison on a real data set of 7,017 cycles collected by 881 women. We demonstrate that the DBN model is more accurate than the best modern Fertility Awareness Methods, based on the observation of mucus, marking reasonably high percentage of days of the cycle as infertile. We argue that the DBN approach offers other advantages, such as predicting the ovulation day and being able to adjust its predictions to each woman's individual cycle