A dual hybrid forecasting model for support of decision making in healthcare management

Forecasting of time series data such as fertility, morbidity and mortality rates is important for healthcare managers as these data serve as health indicators of a society. Accurate forecasting of these data based on past values helps the healthcare managers in taking appropriate decisions for avoiding possible calamity situations. Healthcare time series data consist of complex linear and nonlinear patterns and it may be difficult to obtain high forecasting accuracy rates using only linear or neural network models. In this paper, we present a dual hybrid forecasting model based on soft computing technology. The proposed method makes use of a combination of linear regression, neural network and fuzzy models. The inputs to the fuzzy model are the forecast values of healthcare time series data. Based on a set of rules, the fuzzy model yields a qualitative output which is useful for decision making in healthcare management.

► Accurate forecasting of healthcare data is important for healthcare managers.
► Healthcare data are generally complex comprising linear and nonlinear patterns.
► Dual hybrid model provides more accurate forecasting of healthcare time series data.
► The new method uses both linear regression and neural network models for prediction.
► Fuzzy logic is used to implement action plan which helps healthcare managers.