Comparing the Effects of Cognitive Style, Subjective Emotion,and Physiological Phenomenon on the Accuracy of Intuitive Time-SeriesForecasting Using an SONN: A Proposal

摘要

Self-organizing neural network (SONN) is known tornbe able to extract features in input samples [Kohonen,rn1995]. By updating not only the weight vector of thernwinning neuron in the self-organizing layer but also thosernof its neighboring neurons, neighboring neurons wouldrneventually become to respond similarly to a specific inputrnvector. Then the distribution of winning neurons for arnclass may be distinguished from those for other classes.rnLuttrell proposed a SONN which can inherently use therncorrelation between input vectors of separate clusters andrnhe called it self-supervised adaptive neural networkrn[Luttrell, 1992].rnIn this report, we propose the use of the selfsupervisedrnadaptive algorithm in analyzing the correlationrnbetween cognitive style and the accuracy of intuitiverntime-series forecasting, and suggest a way to compare thernrelative degree of correlation between each of cognitivernstyle, subjective emotion and physiological phenomenonrnand the accuracy of intuitive time-series forecasting.