Application of the BORUTA Algorithm to Input Data Selection for a Model Based on Rough Set Theory (RST) to Prediction Energy Consumption for Building Heating

Energy prediction used for building heating has attracted particular attention because it is often required in the development of various strategies to improve energy efficiency buildings, especially those undergoing thermal improvements. The complexity, dynamics, uncertainty, and nonlinearity existing systems create a great need modeling techniques. One them machine learning models, which are based on input data consisting features that describe objects under study. describing actual buildings build model may be characterized by missing values, duplicate or inconsistent features, noise, outliers. Therefore, an extremely important aspect effort proper selection simplify consumption heating. In this connection, goal was evaluate usefulness final demand rate using groups residential retrofit. created combining two algorithms: BORUTA feature algorithm, prepares conditional variables corresponding rough set theory (RST). research conducted group 109 multi-family from end last century (made large-panel technology), thermomodernized at beginning 21st century. Evaluation metrics such as MAPE, MBE, CV RMSE, R2, adopted statistical calibration standards ASHRAE, were assess quality developed model. analysis obtained results indicated RST, selected gives satisfactory with limited number variables, thus allows predict (after improvement) type high accuracy.