Two methods for linear mapping of multidimensional data in the case of unsupervised learning are proposed. The first method maximizes the mean square density gradient of the projected samples with the intention of com...
详细信息
Two methods for linear mapping of multidimensional data in the case of unsupervised learning are proposed. The first method maximizes the mean square density gradient of the projected samples with the intention of compressing the clusters. The second method is based on the k-NN technique and obtains a map of the scatter of the neighbor clusters. An experiment with the classical Iris data shows the mapping accuracy of the latter method.
This paper describes the face method-a data representation method by faces-applied in the graphical representation and analysis of complex multivariate data. For this, a face pattern similar to a real human face is de...
详细信息
This paper describes the face method-a data representation method by faces-applied in the graphical representation and analysis of complex multivariate data. For this, a face pattern similar to a real human face is designed and constructed. A representation method is proposed to utilize the human visual characteritics of facial expressions, i.e. the representation of respective data groups by typical facial expressions. The efficiencies of the method are tested in plant dynamic data and psychotic data analyses.
The applications of face pattern as a medium to illustrate to man complex computer-processed medical diagnosis are described. A case of nephrotic syndrome was selected to show the mechanics of the study. The design of...
详细信息
The applications of face pattern as a medium to illustrate to man complex computer-processed medical diagnosis are described. A case of nephrotic syndrome was selected to show the mechanics of the study. The design of the face pattern was constructed and psychometric experiments analyzing the resulting facial expressions were conducted. The results obtained from the analysis of facial expressions based and constructed from the original findings of conventional medical methods were used to loosely predict the patient''s future conditions, and the same results were compared with those of statistical procedures to test the efficiency of the method. A separate application of the method was also performed to indicate a possible effect of corticosteroid on idiopathic nephrotic syndrome.
A number of mapping operations in patternrecognition is reviewed. Two families of new mapping algorithms are defined. The performance of all algorithms is illustrated using a collection of 100 feature vectors obtaine...
详细信息
A number of mapping operations in patternrecognition is reviewed. Two families of new mapping algorithms are defined. The performance of all algorithms is illustrated using a collection of 100 feature vectors obtained from images of two classes of white blood cells. It is emphasized that the interactive pattern recognition system ISPAHAN is well suited to find optimal decision functions based on such mappings.
In this paper a general-purpose interactive pattern recognition system is built using limited resources. The system utilizes Ribak's [5] relationship matrix (R-matrix) as a data structure and a PDP-9 minicomputer ...
详细信息
In this paper a general-purpose interactive pattern recognition system is built using limited resources. The system utilizes Ribak's [5] relationship matrix (R-matrix) as a data structure and a PDP-9 minicomputer as hardware. The software developed for the interactive system along with the implemented idea of "vague" relations gives the human operator flexibility and aid in interactive file creation. Files created to describe complex patterns can be formed by merging simpler subfiles of different portions of that pattem. In addition to merging files (union operation), common elements between files can be detected using an intersection operation. User-defined routines are readily interfaced to the interactive system, thus increasing the system flexibility and giving the human operator a tailor-made effect.
This paper presents a new computerized technique to aid the designers of pattern classifiers when the measurement variables are discrete and the values form a simple nominal scale (no inherent metric). A theory of “p...
详细信息
暂无评论