USING NEURAL NETWORK MODELING TO PREDICT THE COURSE OF ACUTE PANCREATITIS

The database of studies of 82 patients with acute pancreatitis are presented. Using neural network analysis, the most indicative parameters for predicting acute pancreatitis were revealed: indexes of Kalf-Kalif intoxication modified by Kostyuchenko and Khomich, Reis, Garkavi, the ratio of leukocytes to ESR, leukocyte index, general intoxication index; sonographic parameters – the size of the head of the pancreas, the diameter of the splenic vein, the presence of free fluid in the abdominal cavity; biochemical parameters – blood amylase concentration, urine diastase. When conducting clustering in a multidimensional feature space, a Kohonen neural network was created. All analyzed objects were effectively divided into 3 clusters. The most severe and prognostically unfavorable is cluster 1, which included data from 30 patients, with the maximum mortality rate and maximum hospital stay.

The Use of the Kohonen Neural Network for Comparing the Declared and Actual State of Knowledge Regarding Reproductive Health and the Impact of Selected Lifestyle Components on Reproductive Health

Infertility is a global problem affecting 48 to 186 million couples of reproductive age. In Poland, it concerns approx. 1.5 million couples, which amounts to 20% of the population capable of reproducing. One of the factors influencing the incidence of fertility disorders may be lifestyle, understood as a multi-disciplinary accumulation of everyday behaviours and habits. In the study, a group of 201 young adults, students of medical and related faculties, were surveyed in order to check the actual level of knowledge about the impact of lifestyle on reproductive health. The Kohonen network, which is an example of a self-learning neural network, was used to find non-obvious connections between the data. The trained Kohonen neural network formed 4 clusters with different characteristics. Based on analyses of the structure of each cluster, it was found that 2nd year students of Medicine are internally divided into 3 fractions. The first fraction declared a high level of knowledge, but did not have real knowledge. The second fraction was aware of their ignorance, as confirmed by the knowledge test. The last fraction was characterized by a high level of self-confidence regarding their knowledge about reproductive health and obtained a high result in the knowledge test. It was confirmed that people studying at the Medical faculty know more than students of the same year at faculties other than Medicine. Interesting results were obtained for a group of 3rd year students of first-cycle studies in Dietetics. They did not obtain a significantly better result in the knowledge test concerning the influence of diet and lifestyle on reproductive health. It would seem that one could expect at least a few highly knowledgeable students in a group of 3rd year students, but this was not confirmed by the study. In view of the obtained results, it was concluded that the Kohonen neural network is applicable to the analysis of data on the actual state of knowledge about the impact of lifestyle on reproductive health.

Application of the Kohonen neural network for monitoring tissue oxygen supply under hypoxic conditions

The article presents the results of application of the Kohonen artificial neural network (KANN) in assessing the oxygen status of human tissues, as well as for studying the adaptive-compensatory response of the body to functional load. In the experiment, the registered digital oxygen images of the tissue of 31 subjects were distributed into three classes using the KANN. Each group is characterised by different resistance of the organism to hypoxia. The research results have shown the effectiveness of using an artificial neural network structure and the possibility of its implementation for recognition of the functional state of a person under conditions of metabolic hypoxia; it seems relevant and has theoretical and practical significance in the framework of ecological physiology.

Network Game of Student Training Quality Assessment System Based on KOHONEN Neural Network

Objectives: In recent years, with the continuous improvement of the requirements of student training quality, the evaluation results of the existing evaluation system of student training quality are mostly unsatisfactory. Therefore, by integrating c-mean algorithm and Kohonen clustering algorithm, a non-sequential artificial neural network is obtained, a student training quality evaluation system based on KOHONEN neural network is designed by automatically adjusting the size of the objective function nodes of the non-sequential artificial neural network. Then the evaluation system is applied to the expected evaluation of the training quality of students in two science classes of Xinghua Middle School in Shenyang, Liaoning Province. The comparison between the test result data and the expected results of the model after the experiment confirms that the evaluation results obtained by using the evaluation system based on KOHONEN neural network have high accuracy.

Kohonen Neural Network Investigation of the Effects of the Visual, Proprioceptive and Vestibular Systems to Balance in Young Healthy Adult Subjects

Kohonen neural network (KNN) was used to investigate the effects of the visual, proprioceptive and vestibular systems using the sway information in the mediolateral (ML) and anterior-posterior (AP) directions, obtained from an inertial measurement unit, placed at the lower backs of 23 healthy adult subjects (10 males, 13 females, mean (standard deviation) age: 24.5 (4.0) years, height: 173.6 (6.8) centimeter, weight: 72.7 (9.9) kg). The measurements were based on the modified Clinical Test of Sensory Interaction and Balance (mCTSIB). KNN clustered the subjects’ time-domain sway measures by processing their sway’s root mean square position, velocity, and acceleration. Clustering effectiveness was established using external performance indicators such as purity, precision-recall, and F-measure. Differences in these measures, from the clustering of each mCTSIB condition with its condition, were used to extract information about the balance-related sensory systems, where smaller values indicated reduced sway differences. The results for the parameters of purity, precision, recall, and F-measure were higher in the AP direction as compared to the ML direction by 7.12%, 11.64%, 7.12%, and 9.50% respectively, with their differences statistically significant (p < 0.05) thus suggesting the related sensory systems affect majorly the AP direction sway as compared to the ML direction sway. Sway differences in the ML direction were lowest in the presence of the visual system. It was concluded that the effect of the visual system on the balance can be examined mostly by the ML sway while the proprioceptive and vestibular systems can be examined mostly by the AP direction sway.

Research on Influence of Attribute Frame Effect on Loan Decision of Undergraduate and Risk Assessment Model of Undergraduate Loan Behavior

The article uses virtual lending scenarios to study the influence of attribute frame effect on undergraduates’ loan decisions. The results show that undergraduates have attribute frame effects in the three major areas of electronic products, life entertainment, and learning and training. There is a significant difference between the positive frame and the negative frame; that is, they are more inclined to make loan decisions under the positive frame. According to the research results, the article designs a loan risk assessment model based on Kohonen neural network and conducts simulation experiments. The experimental results show that the model’ classification accuracy is 72.65%.

Effectiveness Analysis of Rolling Bearing Fault Detectors Based On Self-Organising Kohonen Neural Network – A Case Study of PMSM Drive

Due to their many advantages, permanent magnet synchronous motors (PMSMs) are increasingly used in not only industrial drive systems but also electric and hybrid vehicle drives, aviation and other applications. Unfortunately, PMSMs are not free from damage that occurs during their operation. It is assumed that about 40% of the damage that occurs is related to rolling bearing damage. This article focuses on the use of Kohonen neural network (KNN) for rolling bearing damage detection in a PMSM drive system. The symptoms from the fast Fourier transform (FFT) and Envelope (ENV) Analysis of the mechanical vibration acceleration signal were analysed. The signal ENV was obtained by applying the Hilbert transform (HT). Two neural network functions are discussed: a detector and a classifier. The detector detected the damage and the classifier determined the type of damage to the rolling bearing (undamaged bearing, damaged rolling element, outer or inner race). The effectiveness of the analysed networks from the point of view of the applied signal processing method, map size, type of neighbourhood radius, distance function and the influence of input data normalisation are presented. The results are presented in the form of a confusion matrix, together with 2D and 3D maps of active neurons.