Title :
The reduced Parzen classifier
Author :
Fukunaga, Keinosuke ; Hayes, Raymond R.
Author_Institution :
Sch. of Electr. Eng., Purdue Univ., West Lafayette, IN, USA
fDate :
4/1/1989 12:00:00 AM
Abstract :
The Parzen density estimate is known to be an effective tool for estimating the Bayes error, given a set of training samples from the class distributions. An algorithm is developed to select a given number of representative samples whose Parzen density estimate closely matches that of the entire sample set. Using this reduced representative set, a piecewise quadratic classifier which provides nearly optimal performance is designed.<>
Keywords :
Bayes methods; error analysis; estimation theory; pattern recognition; Bayes error; Parzen classifier; Parzen density estimate; pattern recognition; piecewise quadratic classifier; representative samples; Algorithm design and analysis; Covariance matrix; Design optimization; Error analysis; Gaussian distribution; Kernel; Millimeter wave radar; Pattern recognition; Probability distribution; Shape;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
