Computational limitations of affine automata and generalized affine automata

Abstract We present new results on the computational limitations of affine automata (AfAs). First, we show that using endmarker does not increase power AfAs. Second, computation bounded-error rational-valued AfAs can be simulated in logarithmic space. Third, identify some logspace unary languages are recognized by algebraic-valued Fourth, arbitrary real-valued transition matrices and state vectors unbounded-error model. When focusing only rational values, obtain same result also for bounded error. As a consequence, class remains when restricted to use numbers only.