Affective Autonomous Agents For Supporting Investment Decision Processes Using Artificial Somatic Reactions

Sometimes, the conscious act of decision-making in humans is dramatically interrupted by situations that warrant an immediate response (e.g. when there is an imminent risk). The human body somatizes this interruption such that an action could be taken without a rational analysis. The above is known as a somatic marker. According to the somatic marker hypothesis, somatic markers could directly influence several ambits of decision-making. This research work presents the incorporation of artificial somatic reactions into affective autonomous agents who implement decision-making in the stock market. This implies the design of a general decision architecture for stock markets considering artificial somatic reactions and the definition of a set of decision-making algorithms for supporting investment decisions performed by affective autonomous agents (considering artificial somatic reactions). Test scenarios were defined using official data from Standard & Poor's 500 and Dow Jones. The experimental results are promising and indicated that affective autonomous agents are able to experience artificial somatic reactions and achieve effectiveness and efficiency in their decision-making.