Predicting Emotions from Temporal Physical and Behavioral Information
Open Access
- Graduate Program:
- Industrial Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- July 27, 2015
- Committee Members:
- Conrad S Tucker, Thesis Advisor
- Keywords:
- Emotion Prediction
- Forecasting
- Machine Learning
- Abstract:
- It is believed that mental anomalies such as stress and anxiety not only cause suffering for the individuals, but also lead to tragedies in some extreme cases. The ability to predict the mental states of an individual could prove critical to healthcare practitioners. Currently, the practical ways to predict an individual's mental state is through mental examinations that involve psychological experts performing the evaluations. However, such methods may be time and resource consuming, mitigating its broad applicability to a wide population. Furthermore, some individuals may also be unaware of their mental anomalies or may not feel uncomfortable to express themselves during the evaluations. Hence, their mental anomalies could remain undetected for a prolonged period of time. The hypothesis of this work is to prove that current and future mental states of an individual could be estimated using only the information directly observable from the his/her physical activities and behaviors. The problem of emotion prediction is transformed into the time series forecasting problem where an individual is represented as a multivariate time series stream of monitored physical and behavioral attributes. A mathematical model is then generated to capture the dependencies among these attributes, which is used for prediction of emotion states. In particular, we first illustrate the drawbacks of traditional multivariate time series forecasting methodology such as vector autoregression. Then, we show that such issues could be mitigated by using machine learning regression techniques which are modified for capturing temporal dependencies in multivariate time series data. A case study using the data from 150 human participants from the Pennsylvania State University community illustrates that the proposed machine learning based forecasting methods are more suitable for high-dimensional psychological data than the traditional vector autoregressive model in terms of both magnitude of error and directional accuracy. These results not only present a successful usage of machine learning techniques in psychological studies, but also serve as a building block for multiple medical applications that could rely on an automated system to gauge individuals' states of emotion.