Depression in the elderly is an important social issue considering the population aging of the world. In particular, elderly living alone who has narrowed social relationship due to bereavement and retirement are more prone to be depressed. Long-term depressed mood can be a precursor to eventual depression as a disease. Our goal is how to predict the depressed mood of single household elderly from unobtrusive monitoring of their daily life. We have selected a wearable band with multiple sensors for monitoring elderly people. Depression questionnaire has been surveyed periodically to be used as the labels. Instead of working with depression patients, we recruited 14 single household elderly people from a nearby community. The wearable band provided daily activity and biometric data for 71 days. From the data, we generate a depressed mood prediction model. Multiple features from the collected sensor data are exploited for model generation. One general model is generated to be used as the baseline for the initial model deployment. Personal models are also generated for model refinement. The general model has a high recall of 80% in an MLP model. Individual models achieved an average recall of 82.7%. In this study, we have demonstrated that we can generate depressed mood prediction models with data collected from real daily living. Our work has shown the feasibility of using a wearable band as an unobtrusive depression monitoring sensor even for elderly people.

Virtual metrology (VM) has been successfully applied to semiconductor manufacturing as an efficient way of achieving wafer-to-wafer quality control. VM involves the estimation of metrology variables of wafer inspection using a prediction model trained with process parameters and measurements prior to the actual implementation of metrology. VM modeling should incorporate a number of process parameters and measurements collected from each piece of process equipment, which results in a greater number of input variables. Therefore, it is necessary to resolve the problem of high dimensionality through feature selection. A suitable feature selection method for VM modeling should effectively address the high dimensionality by lowering the computational cost, while also achieving high prediction accuracy as an essential requirement for the practical deployment of VM. In this paper, a feature selection method based on random forward search is proposed for efficient VM modeling. This method selects relevant variables sequentially from disjoint random subsets of candidate variables by incorporating randomness. Our preliminary experimental results obtained with real-world semiconductor manufacturing data demonstrate that the proposed feature selection method achieves comparable prediction accuracy yet has the advantages of being computationally more efficient, thus merits further investigation.