G. Deichsel, H.J. Trampisch "Clusteranalyse und Diskriminanzanalyse" Gustav Fisher Verlag, Stuttgart, 1985. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Integration of Neural Networks with Knowledge-Based Systems - Ultsch, Korus (1995) (3 citations) (Correct)
....class borders, labeling of clusters, and in addition, single component maps, which show the distribution of a single feature on the SOFM. For example, using a data set containing blood analysis values from 20 patients (20 vectors with 11 real valued components) selected from a set of 1500 patients [3], it turned out that the clustering coresponds nicely with the different patient s diagnoses. 4. Integrated Knowledge Acquisition In the previous Section we presented the combination of the Self organization Feature Map (SOFM) by Kohonen [5] and the U matrix methods [19] to detect structure in ....
G. Deichsel, H.J. Trampisch "Clusteranalyse und Diskriminanzanalyse" Gustav Fisher Verlag, Stuttgart, 1985.
Concept Formation by Combining Neural Networks and Machine Learning - Sklorz (Correct)
.... x k generate clusters in the input space. Figure 3 a) visualizes this thought. E C C D B B A B B B A A A A A A D D A C Figure 3. a) Distribution of cluster centers b) P matrix Following this idea, our realization of SOFM was trained 2 with a medical data set introduced by [7]. This data was drawn from a body of blood measurements. Each of the 20 object vectors consisted of 11 different blood values 3 . According to medical diagnosis, the object vectors form 5 clusters (A: 8 normal; B: 5 respiratory acidosis; C: 3 lactacidosis; D: 3 metabolical acidosis; E: 1 ....
....the values of the parameters are randomly chosen from the ranges: t m 2 f700; 701; 1200g; t c 2 f10; 11; 25g; ff m 2 [0:07; 0:14] ff c 2 [0:1; 0:2] net 2 2 f900; 1600; 2500g. We briefly present further experiments with the following two medical data sets. ffl Data set 1 [7] consists of 20 object vectors. Each object vector is described by two attributes: hemoglobin (FE, mg l) serum levels of alkaline phosphatase (AP, U l) According to medical diagnosis, the object vectors form 4 clusters: A: 5 normal; B: 6 hepatitis; C: 5 obstructive jaundice; D: 4 liver ....
G. Deichsel and H.J. Trampisch. Clusteranalyse und Diskriminanzanalyse. Gustav Fischer Verlag, Stuttgard, Germany, 1985. in German.
Integration of Neural Networks and Knowledge-Based Systems in.. - Ultsch, al. (1995) (1 citation) (Correct)
....1995, pp. 425 426. Application in Medicine In order to test our hybrid system we applied it to two medical applications. First, we used it to diagnose acidosis diseases. The data set consists of 11 attributes originating from the blood analysis. Several classification methods according to [Deichsel Trampisch85] were used to explain these data. The Neural Network together with the UMatrix method was able to classify the data into the subcategories healthy, lacacidemia, metabolical acidosis, respiratory acidosis and one patient with cerebral deficiency (Fig.1a) With our rule generation module Fig. 1 ....
Deichsel, G.; Trampisch, H.J.: Clusteranalyse und Diskriminanzanalyse, Fischer 1985.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC