About the new algoritm of characterization of $k$-valued threshold functions

The article provides an overview of the known approaches to the characterization (learning) of Boolean and $k$-valued threshold functions. Proposed a new algorithm characterization of $k$-valued threshold functions for which we use expansion coefficients and increase coefficients for initial approximation of the coefficients of the linear form. Also we give the results of experimental comparisons of the new algorithm with a known Obradovic learning algorithm.