Feature selection, as a preprocessing step to machine learning, has been effective in reducing dimensionality, removing irrelevant data, increasing learning accuracy, and improvin...
Feature selection is an important task in effective data mining. A new challenge to feature selection is the so-called “small labeled-sample problem” in which labeled data is...
We report on a property-based approach to feature interaction analysis for a client-server email system. The model is based upon Hall’s email model [12] presented at FIW’00 [3]...
In nature, animals encounter high dimensional sensory stimuli that have complex statistical and dynamical structure. Attempts to study the neural coding of these natural signals f...
B. D. Wright, Kamal Sen, William Bialek, A. J. Dou...
An automatic recognition of online handwritten text has been an on-going research problem for nearly four decades. It has been gaining more interest due to the increasing populari...