Parameter learning of bayesian network classifiers under computational constraints

Abstract

We consider online learning of Bayesian network classifiers (BNCs) with reduced-precision parameters, i.e. the conditional-probability tables parameterizing the BNCs are represented by low bit-width fixed-point numbers. In contrast to previous work, we analyze the learning of these parameters using reduced-precision arithmetic only which is important for computationally constrained platforms, e.g. embeddedand ambient-systems, as well as power-aware systems. This requires specialized algorithms since naive implementations of the projection for ensuring the sum-to-one constraint of the parameters in gradient-based learning are not sufficiently accurate. In particular, we present generative and discriminative learning algorithms for BNCs relying only on reduced-precision arithmetic. For several standard benchmark datasets, these algorithms achieve classification-rate performance close to that of BNCs with parameters learned by conventional algorithms using doubleprecision floating-point arithmetic. Our results facilitate the utilization of BNCs in the foresaid systems.