Decision Making guided by Emotion A computational architecture
This work addresses decision-making in AI or cognitive modeling, but it appears incremental as it builds on existing neurobiological studies and Guided Propagation Networks without clear broad impact.
The paper tackles the problem of integrating emotional and cognitive processes in decision-making by proposing a computational architecture where emotional channels guide a decision-making channel, and reports simulation results showing the integrated contribution of emotional influences and issues with accidental all-out emotions.
A computational architecture is presented, in which "swift and fuzzy" emotional channels guide a "slow and precise" decision-making channel. Reported neurobiological studies first provide hints on the representation of both emotional and cognitive dimensions across brain structures, mediated by the neuromodulation system. The related model is based on Guided Propagation Networks, the inner flows of which can be guided through modulation. A key-channel of this model grows from a few emotional cues, and is aimed at anticipating the consequences of ongoing possible actions. Current experimental results of a computer simulation show the integrated contribution of several emotional influences, as well as issues of accidental all-out emotions.