Pattern Recognition using Optimized Prototypes
نویسنده
چکیده
Hong Yan Department of Electrical Engineering, University of Sydney, NSW 2006, Australia Phone: +61 2 692-3515, fax: +61 2 692-3847, e-mail: [email protected] Abstract In this paper we present a method for designing optimized nearest neighbor classi ers. In this method, we produce initial prototypes for classi cation based on self-organization and optimize the prototypes using a multi-layer neural network. Since only a very small number of prototypes are needed for classi cation when they are optimized, the classi er can be implemented very e ciently.
منابع مشابه
Chronos-trainer: a Rapid Prototyping Framework for Realtime Audio Pattern Recognition
We present the Chronos-Trainer Framework for developing audio pattern recognition systems with a focus on both offline prototyping and realtime execution. The framework covers feature extraction, labelling, classifier training, evaluation and execution. It is implemented as a hybrid Simulink MATLAB C++ Java Framework which enables seamless transformation of algorithms from MATLABand Simulink-pr...
متن کاملInvestigation of the effective factors on manufacturing calcium phosphates prototypes using 3D printing
Calcium phosphate ceramics has been widely used in the present due to their chemical similarity to bone and good biocompatibility in the physiological environmental and a compatibility with synthetic and natural polymers Recent advancements in additive manufacturing have enabled the fabrication of 3D prototypes with controlled architecture resembling the natural bone. Binder jetting is a versat...
متن کاملFilter-based models for pattern classification
-In this paper we consider a technique for pattern classification based upon the development of prototypes which capture the distinguishing features ("disjunctive prototypes") of each pattern class and, via cross-correlation with incoming test images, enable efficient pattern classification. We evaluate such a classification procedure with prototypes based on the images per se (direct code), Ga...
متن کاملOptimization of Concept Prototypes for the Recognition of Noisy Texture Data
This paper presents an approach to the recognition of noisy vision concepts incorporating machine learning and concept optimization techniques. Noisy texture characteristics require the development of techniques that can improve the performance of a texture recognition system. We develop such techniques through the optimization of learned concept prototypes in order to remove noisy and less sig...
متن کاملA prototype classification method and its use in a hybrid solution for multiclass pattern recognition
In order to combine a fast multiclass classification method with an effective binary classification method, we have developed a prototype learning/matching scheme that can be integrated with support vector machines (SVM) for vector-matching applications. This prototype classification method employs a learning process to determine both the number and the location of prototypes. The learning proc...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1994