This paper shows that the nonlinear relationships between the personal parameters (age, gender experiences, physical conditions etc.) and the outputs of POMS (Profile of Moods State) score could be identified by using a neural network. The neural network model can noninvasively measure the POMS scores needed in order to develop a psychological response feedback control system for stress management. While the research subjects were engaged in horticultural activity, their age, gender, activity time, interest in gardening and the contents of their activities were examined in relation to the change in their POMS scores in a psychological test. The results suggested that POMS change could be investigated through stress index parameters.
The neural network model developed was tested with 10 trial subjects. The error obtained was less than 10%. The systems approach discussed in the research can be applied in the field of horticultural therapy and nonlinear modeling tools such as neural networks can identify the nonlinear systems involved in such systems.
View full abstract