The household energy consumption has a large share of global energy consumption. To have better understanding of energy generation, management and surplus storage, we need to discover implicit patterns of consumers’ behavior and identify the factors affecting their performance. The main goal of this paper is to descriptively analyze the pattern of household energy consumption using RECS2015 dataset. To this end, we focus on selecting the most effective subset of features from high dimensional dataset that leads to a better understanding of data, reducing computation time and improving prediction performance. The result of this study can help decision makers to investigate the living conditions of families in different levels of society to ensure that their life style is well enough or should be improved.