A Multialternative Neural Decision Process

We introduce an algorithmic decision process for multialternative choice that combines binary comparisons and Markovian exploration. We show that a functional property, transitivity, makes it testable.