福州大学智能技术课件-10-FeatureSelection.pdfVIP

Brief Introduction to Feature Selection/Extraction 2017-6-15 智能技术课程D-Y YE In practice, a set of data may contain a large number of original features that are either redundant or irrelevant to classification. Training a model out of such data is often computationally intractable due to curse of dimensionality, and what is worse, the trained model is usually of poor generalization ability with large variance or over-fitting problem. Therefore, a dimensionality reduction process is necessary before training a model or clustering dat