Abstract
The metabolic disorder osteoporosis has affected a humongous number of individuals globally. Its progression can be slowed down by modifying lifestyle risk factors and by following appropriate treatment. In this research work, modifiable risk factors of osteoporosis have been considered. All these variables are binary thus providing incomplete information. Machine learning implementation on these factors took a large computation time and has shown poor accuracy. Thus fuzzy concept has been introduced leading to the development of a fusion of machine learning and fuzzy logic approach. Three binary variables of the considered dataset have been compared thus fuzzy input is produced which also considers the uncertainty of these binary variables and since three input variables are transformed into one the number of features has also been reduced leading to optimization of computation time and accuracy. Moreover, it guides the individual to modify lifestyle factors to slow down the disease progression or reduce the risk of osteoporosis. The proposed model is validated on the diabetes risk prediction dataset.•The study examines modifiable binary risk factors for osteoporosis, such as diet, smoking, and exercise etc.•A fusion of machine learning and fuzzy logic is introduced to improve accuracy and reduce computation time.•The model, which condenses three binary inputs into one, is validated using a diabetes risk prediction dataset.