States evolution in Θ(λ)-learning based on logical MDPs with negation | IEEE Conference Publication | IEEE Xplore