Catching the recent trend of data is an important issue when mining frequent itemsets from data streams. To prevent from storing the whole transaction data within the sliding windo...
We model on-line ink traces for a set of 219 symbols to “best fit” low-degree polynomial series. Using a collection of mathematical writing samples, we find that in many cas...
We propose a new language learning model that learns a syntactic-semantic grammar from a small number of natural language strings annotated with their semantics, along with basic ...