A New Kernel to use with Discretized Temporal Series

In this paper a new Kernel, from statistical learning theory is proposed to work with symbols chains (words) obtained from a discretization procedure of a continuous features. Although the exact definition of the discretization is not strictly necessary, there must always exist either, a measure of distance or a similarity between symbols in a certain alphabet (a set of symbols). This kernel is applied on a set of television shares obtained from the seven main television stations in Andalusia (Spain). A comparative study for classification purposes is done, and the associated parameter selection is studied. Finally, it must be mentioned that this kernel has certain implications in the type of considered similarity that will be studied in further researches. The small influence of the λ parameter in identification tasks must also be discussed.