Introduction Reinforcement learning (RL) has been shown to be an effective paradigm for learning control policies for problems with discrete state spaces. For problems with continuous multi-dimensional state spaces, the results are less compelling. When these state spaces can be effectively discretized, traditional techniques can be applied. However, many interesting problems must be discretize...

In this paper we consider a construction in an arbitrary triangulated category T which resembles the notion of a Moore spectrum in algebraic topology. Namely, given a compact object C of T satisfying some finite tilting assumptions, we obtain a functor which “approximates” objects from the module category of the endomorphism algebra of C in T . This generalises and extends a construction of Jør...