Daily activity forecasts play an important role in the daily lives of residents in smart homes. Category forecasts and occurrence time forecasts of daily activity are two key tasks. Category forecasts of daily activity are correlated with occurrence time forecasts, however, existing research has only focused on one of the two tasks. Moreover, the performance of daily activity forecasts is low when the two tasks are performed in series. In this paper, a forecast model based on multi-task learning is proposed to forecast category and occurrence time of daily activity mutually and iteratively. Firstly, raw sensor events are pre-processed to form a feature space of daily activity. Secondly, a parallel multi-task learning model which combines a convolutional neural network (CNN) with bidirectional long short-term memory (Bi-LSTM) units are developed as the forecast model. Finally, five distinct datasets are used to evaluate the proposed model. The experimental results show that compared with the state-of-the-art single-task learning models, this model improves accuracy by at least 2.22%, and the metrics of NMAE, NRMSE and R² are improved by at least 1.542%, 7.79% and 1.69%, respectively.

Activity recognition benefits the lives of residents in a smart home on a daily basis. One of the aims of this technology is to achieve good performance in activity recognition. The extraction and selection of the daily activity feature have a significant effect on this performance. However, commonly used extraction of daily activity features have limited the performance of daily activity recognition. Based on the nature of the time series of sensor events caused by daily activities, this paper presents a novel extraction approach for daily activity feature. First, time tuples are extracted from sensor events to form a time series. Subsequently, several common statistic formulas are proposed to form the space of daily activity features. Finally, a feature selection algorithm is employed to generate final daily activity features. To evaluate the proposed approach, two distinct datasets are adopted for activity recognition based on four different classifiers. The results of the experiment reveal that the proposed approach is an improvement over the commonly used approach.