Attribute-efficient learning functions of k-valued logic

An attribute-efficient learning linear functions of k-valued logic in exact learning model is considered in this work. Here we obtain the upper and lower bounds for two types of queries: value queries and comparative queries. Also the complexity of learning linear boolean functions of 2 or 3 relevant variables was obtained. The order of complexity of learning functions was obtained for both types of queries when the number of variables tends to infinity.