• Korean Journal of Radiology
• /
• v.23 no.10
• /
• pp.949-958
• /
• 2022

Objective: To investigate the feasibility of using a deep learning-based analysis of auscultation data to predict significant stenosis of arteriovenous fistulas (AVF) in patients undergoing hemodialysis requiring percutaneous transluminal angioplasty (PTA). Materials and Methods: Forty patients (24 male and 16 female; median age, 62.5 years) with dysfunctional native AVF were prospectively recruited. Digital sounds from the AVF shunt were recorded using a wireless electronic stethoscope before (pre-PTA) and after PTA (post-PTA), and the audio files were subsequently converted to mel spectrograms, which were used to construct various deep convolutional neural network (DCNN) models (DenseNet201, EfficientNetB5, and ResNet50). The performance of these models for diagnosing ≥ 50% AVF stenosis was assessed and compared. The ground truth for the presence of ≥ 50% AVF stenosis was obtained using digital subtraction angiography. Gradient-weighted class activation mapping (Grad-CAM) was used to produce visual explanations for DCNN model decisions. Results: Eighty audio files were obtained from the 40 recruited patients and pooled for the study. Mel spectrograms of "pre-PTA" shunt sounds showed patterns corresponding to abnormal high-pitched bruits with systolic accentuation observed in patients with stenotic AVF. The ResNet50 and EfficientNetB5 models yielded an area under the receiver operating characteristic curve of 0.99 and 0.98, respectively, at optimized epochs for predicting ≥ 50% AVF stenosis. However, Grad-CAM heatmaps revealed that only ResNet50 highlighted areas relevant to AVF stenosis in the mel spectrogram. Conclusion: Mel spectrogram-based DCNN models, particularly ResNet50, successfully predicted the presence of significant AVF stenosis requiring PTA in this feasibility study and may potentially be used in AVF surveillance.

• Journal of Bio-Environment Control
• /
• v.28 no.2
• /
• pp.95-103
• /
• 2019

Temperature and relative humidity are important factors in crop cultivation and should be properly controlled for improving crop yield and quality. In order to control the environment accurately, we need to predict how the environment will change in the future. The objective of this study was to predict air temperature and relative humidity at a future time by using a multilayer perceptron (MLP). The data required to train MLP was collected every 10 min from Oct. 1, 2016 to Feb. 28, 2018 in an eight-span greenhouse ($1,032m^2$) cultivating mango (Mangifera indica cv. Irwin). The inputs for the MLP were greenhouse inside and outside environment data, and set-up and operating values of environment control devices. By using these data, the MLP was trained to predict the air temperature and relative humidity at a future time of 10 to 120 min. Considering typical four seasons in Korea, three-day data of the each season were compared as test data. The MLP was optimized with four hidden layers and 128 nodes for air temperature ($R^2=0.988$) and with four hidden layers and 64 nodes for relative humidity ($R^2=0.990$). Due to the characteristics of MLP, the accuracy decreased as the prediction time became longer. However, air temperature and relative humidity were properly predicted regardless of the environmental changes varied from season to season. For specific data such as spray irrigation, however, the numbers of trained data were too small, resulting in poor predictive accuracy. In this study, air temperature and relative humidity were appropriately predicted through optimization of MLP, but were limited to the experimental greenhouse. Therefore, it is necessary to collect more data from greenhouses at various places and modify the structure of neural network for generalization.

• Information Systems Review
• /
• v.16 no.3
• /
• pp.161-177
• /
• 2014

Corporate credit rating assessment consists of complicated processes in which various factors describing a company are taken into consideration. Such assessment is known to be very expensive since domain experts should be employed to assess the ratings. As a result, the data-driven corporate credit rating prediction using statistical and artificial intelligence (AI) techniques has received considerable attention from researchers and practitioners. In particular, statistical methods such as multiple discriminant analysis (MDA) and multinomial logistic regression analysis (MLOGIT), and AI methods including case-based reasoning (CBR), artificial neural network (ANN), and multiclass support vector machine (MSVM) have been applied to corporate credit rating.2) Among them, MSVM has recently become popular because of its robustness and high prediction accuracy. In this study, we propose a novel optimized MSVM model, and appy it to corporate credit rating prediction in order to enhance the accuracy. Our model, named 'GAMSVM (Genetic Algorithm-optimized Multiclass Support Vector Machine),' is designed to simultaneously optimize the kernel parameters and the feature subset selection. Prior studies like Lorena and de Carvalho (2008), and Chatterjee (2013) show that proper kernel parameters may improve the performance of MSVMs. Also, the results from the studies such as Shieh and Yang (2008) and Chatterjee (2013) imply that appropriate feature selection may lead to higher prediction accuracy. Based on these prior studies, we propose to apply GAMSVM to corporate credit rating prediction. As a tool for optimizing the kernel parameters and the feature subset selection, we suggest genetic algorithm (GA). GA is known as an efficient and effective search method that attempts to simulate the biological evolution phenomenon. By applying genetic operations such as selection, crossover, and mutation, it is designed to gradually improve the search results. Especially, mutation operator prevents GA from falling into the local optima, thus we can find the globally optimal or near-optimal solution using it. GA has popularly been applied to search optimal parameters or feature subset selections of AI techniques including MSVM. With these reasons, we also adopt GA as an optimization tool. To empirically validate the usefulness of GAMSVM, we applied it to a real-world case of credit rating in Korea. Our application is in bond rating, which is the most frequently studied area of credit rating for specific debt issues or other financial obligations. The experimental dataset was collected from a large credit rating company in South Korea. It contained 39 financial ratios of 1,295 companies in the manufacturing industry, and their credit ratings. Using various statistical methods including the one-way ANOVA and the stepwise MDA, we selected 14 financial ratios as the candidate independent variables. The dependent variable, i.e. credit rating, was labeled as four classes: 1(A1); 2(A2); 3(A3); 4(B and C). 80 percent of total data for each class was used for training, and remaining 20 percent was used for validation. And, to overcome small sample size, we applied five-fold cross validation to our dataset. In order to examine the competitiveness of the proposed model, we also experimented several comparative models including MDA, MLOGIT, CBR, ANN and MSVM. In case of MSVM, we adopted One-Against-One (OAO) and DAGSVM (Directed Acyclic Graph SVM) approaches because they are known to be the most accurate approaches among various MSVM approaches. GAMSVM was implemented using LIBSVM-an open-source software, and Evolver 5.5-a commercial software enables GA. Other comparative models were experimented using various statistical and AI packages such as SPSS for Windows, Neuroshell, and Microsoft Excel VBA (Visual Basic for Applications). Experimental results showed that the proposed model-GAMSVM-outperformed all the competitive models. In addition, the model was found to use less independent variables, but to show higher accuracy. In our experiments, five variables such as X7 (total debt), X9 (sales per employee), X13 (years after founded), X15 (accumulated earning to total asset), and X39 (the index related to the cash flows from operating activity) were found to be the most important factors in predicting the corporate credit ratings. However, the values of the finally selected kernel parameters were found to be almost same among the data subsets. To examine whether the predictive performance of GAMSVM was significantly greater than those of other models, we used the McNemar test. As a result, we found that GAMSVM was better than MDA, MLOGIT, CBR, and ANN at the 1% significance level, and better than OAO and DAGSVM at the 5% significance level.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_1
• /
• pp.715-727
• /
• 2023

To provide detailed and appropriate meteorological information in mountainous areas, the Korea Forest Service has established an Automatic Mountain Meteorology Observation Station (AMOS) network in major mountainous regions since 2012, and 464 stations are currently operated. In this study, we proposed an optimal kriging technique with lapse rate correction to produce gridded temperature data suitable for Korean forests using AMOS point observations. First, the outliers of the AMOS temperature data were removed through statistical processing. Then, an optimized theoretical variogram, which best approximates the empirical variogram, was derived to perform the optimal kriging with lapse rate correction. A 500-meter resolution Kriging map for temperature was created to reflect the elevation variations in Korean mountainous terrain. A blind evaluation of the method using a spatially unbiased validation sample showed a correlation coefficient of 0.899 to 0.953 and an error of 0.933 to 1.230℃, indicating a slight accuracy improvement compared to regular kriging without lapse rate correction. However, the critical advantage of the proposed method is that it can appropriately represent the complex terrain of Korean forests, such as local variations in mountainous areas and coastal forests in Gangwon province and topographical differences in Jirisan and Naejangsan and their surrounding forests.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.1-21
• /
• 2020

With the rapid development of artificial intelligence technology, various techniques have been developed to extract meaningful information from unstructured text data which constitutes a large portion of big data. Over the past decades, text mining technologies have been utilized in various industries for practical applications. In the field of business intelligence, it has been employed to discover new market and/or technology opportunities and support rational decision making of business participants. The market information such as market size, market growth rate, and market share is essential for setting companies' business strategies. There has been a continuous demand in various fields for specific product level-market information. However, the information has been generally provided at industry level or broad categories based on classification standards, making it difficult to obtain specific and proper information. In this regard, we propose a new methodology that can estimate the market sizes of product groups at more detailed levels than that of previously offered. We applied Word2Vec algorithm, a neural network based semantic word embedding model, to enable automatic market size estimation from individual companies' product information in a bottom-up manner. The overall process is as follows: First, the data related to product information is collected, refined, and restructured into suitable form for applying Word2Vec model. Next, the preprocessed data is embedded into vector space by Word2Vec and then the product groups are derived by extracting similar products names based on cosine similarity calculation. Finally, the sales data on the extracted products is summated to estimate the market size of the product groups. As an experimental data, text data of product names from Statistics Korea's microdata (345,103 cases) were mapped in multidimensional vector space by Word2Vec training. We performed parameters optimization for training and then applied vector dimension of 300 and window size of 15 as optimized parameters for further experiments. We employed index words of Korean Standard Industry Classification (KSIC) as a product name dataset to more efficiently cluster product groups. The product names which are similar to KSIC indexes were extracted based on cosine similarity. The market size of extracted products as one product category was calculated from individual companies' sales data. The market sizes of 11,654 specific product lines were automatically estimated by the proposed model. For the performance verification, the results were compared with actual market size of some items. The Pearson's correlation coefficient was 0.513. Our approach has several advantages differing from the previous studies. First, text mining and machine learning techniques were applied for the first time on market size estimation, overcoming the limitations of traditional sampling based- or multiple assumption required-methods. In addition, the level of market category can be easily and efficiently adjusted according to the purpose of information use by changing cosine similarity threshold. Furthermore, it has a high potential of practical applications since it can resolve unmet needs for detailed market size information in public and private sectors. Specifically, it can be utilized in technology evaluation and technology commercialization support program conducted by governmental institutions, as well as business strategies consulting and market analysis report publishing by private firms. The limitation of our study is that the presented model needs to be improved in terms of accuracy and reliability. The semantic-based word embedding module can be advanced by giving a proper order in the preprocessed dataset or by combining another algorithm such as Jaccard similarity with Word2Vec. Also, the methods of product group clustering can be changed to other types of unsupervised machine learning algorithm. Our group is currently working on subsequent studies and we expect that it can further improve the performance of the conceptually proposed basic model in this study.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.1041-1043
• /
• 1993

Fuzzy logic based Control Theory has gained much interest in the industrial world, thanks to its ability to formalize and solve in a very natural way many problems that are very difficult to quantify at an analytical level. This paper shows a solution for treating membership function inside hardware circuits. The proposed hardware structure optimizes the memoried size by using particular form of the vectorial representation. The process of memorizing fuzzy sets, i.e. their membership function, has always been one of the more problematic issues for the hardware implementation, due to the quite large memory space that is needed. To simplify such an implementation, it is commonly [1,2,8,9,10,11] used to limit the membership functions either to those having triangular or trapezoidal shape, or pre-definite shape. These kinds of functions are able to cover a large spectrum of applications with a limited usage of memory, since they can be memorized by specifying very few parameters ( ight, base, critical points, etc.). This however results in a loss of computational power due to computation on the medium points. A solution to this problem is obtained by discretizing the universe of discourse U, i.e. by fixing a finite number of points and memorizing the value of the membership functions on such points [3,10,14,15]. Such a solution provides a satisfying computational speed, a very high precision of definitions and gives the users the opportunity to choose membership functions of any shape. However, a significant memory waste can as well be registered. It is indeed possible that for each of the given fuzzy sets many elements of the universe of discourse have a membership value equal to zero. It has also been noticed that almost in all cases common points among fuzzy sets, i.e. points with non null membership values are very few. More specifically, in many applications, for each element u of U, there exists at most three fuzzy sets for which the membership value is ot null [3,5,6,7,12,13]. Our proposal is based on such hypotheses. Moreover, we use a technique that even though it does not restrict the shapes of membership functions, it reduces strongly the computational time for the membership values and optimizes the function memorization. In figure 1 it is represented a term set whose characteristics are common for fuzzy controllers and to which we will refer in the following. The above term set has a universe of discourse with 128 elements (so to have a good resolution), 8 fuzzy sets that describe the term set, 32 levels of discretization for the membership values. Clearly, the number of bits necessary for the given specifications are 5 for 32 truth levels, 3 for 8 membership functions and 7 for 128 levels of resolution. The memory depth is given by the dimension of the universe of the discourse (128 in our case) and it will be represented by the memory rows. The length of a world of memory is defined by: Length = nem (dm(m)＋dm(fm) Where: fm is the maximum number of non null values in every element of the universe of the discourse, dm(m) is the dimension of the values of the membership function m, dm(fm) is the dimension of the word to represent the index of the highest membership function. In our case then Length=24. The memory dimension is therefore 128*24 bits. If we had chosen to memorize all values of the membership functions we would have needed to memorize on each memory row the membership value of each element. Fuzzy sets word dimension is 8*5 bits. Therefore, the dimension of the memory would have been 128*40 bits. Coherently with our hypothesis, in fig. 1 each element of universe of the discourse has a non null membership value on at most three fuzzy sets. Focusing on the elements 32,64,96 of the universe of discourse, they will be memorized as follows: The computation of the rule weights is done by comparing those bits that represent the index of the membership function, with the word of the program memor . The output bus of the Program Memory (μCOD), is given as input a comparator (Combinatory Net). If the index is equal to the bus value then one of the non null weight derives from the rule and it is produced as output, otherwise the output is zero (fig. 2). It is clear, that the memory dimension of the antecedent is in this way reduced since only non null values are memorized. Moreover, the time performance of the system is equivalent to the performance of a system using vectorial memorization of all weights. The dimensioning of the word is influenced by some parameters of the input variable. The most important parameter is the maximum number membership functions (nfm) having a non null value in each element of the universe of discourse. From our study in the field of fuzzy system, we see that typically nfm 3 and there are at most 16 membership function. At any rate, such a value can be increased up to the physical dimensional limit of the antecedent memory. A less important role n the optimization process of the word dimension is played by the number of membership functions defined for each linguistic term. The table below shows the request word dimension as a function of such parameters and compares our proposed method with the method of vectorial memorization[10]. Summing up, the characteristics of our method are: Users are not restricted to membership functions with specific shapes. The number of the fuzzy sets and the resolution of the vertical axis have a very small influence in increasing memory space. Weight computations are done by combinatorial network and therefore the time performance of the system is equivalent to the one of the vectorial method. The number of non null membership values on any element of the universe of discourse is limited. Such a constraint is usually non very restrictive since many controllers obtain a good precision with only three non null weights. The method here briefly described has been adopted by our group in the design of an optimized version of the coprocessor described in [10].

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.125-140
• /
• 2013

We call a data set in which the number of records belonging to a certain class far outnumbers the number of records belonging to the other class, 'imbalanced data set'. Most of the classification techniques perform poorly on imbalanced data sets. When we evaluate the performance of a certain classification technique, we need to measure not only 'accuracy' but also 'sensitivity' and 'specificity'. In a customer churn prediction problem, 'retention' records account for the majority class, and 'churn' records account for the minority class. Sensitivity measures the proportion of actual retentions which are correctly identified as such. Specificity measures the proportion of churns which are correctly identified as such. The poor performance of the classification techniques on imbalanced data sets is due to the low value of specificity. Many previous researches on imbalanced data sets employed 'oversampling' technique where members of the minority class are sampled more than those of the majority class in order to make a relatively balanced data set. When a classification model is constructed using this oversampled balanced data set, specificity can be improved but sensitivity will be decreased. In this research, we developed a hybrid model of support vector machine (SVM), artificial neural network (ANN) and decision tree, that improves specificity while maintaining sensitivity. We named this hybrid model 'hybrid SVM model.' The process of construction and prediction of our hybrid SVM model is as follows. By oversampling from the original imbalanced data set, a balanced data set is prepared. SVM_I model and ANN_I model are constructed using the imbalanced data set, and SVM_B model is constructed using the balanced data set. SVM_I model is superior in sensitivity and SVM_B model is superior in specificity. For a record on which both SVM_I model and SVM_B model make the same prediction, that prediction becomes the final solution. If they make different prediction, the final solution is determined by the discrimination rules obtained by ANN and decision tree. For a record on which SVM_I model and SVM_B model make different predictions, a decision tree model is constructed using ANN_I output value as input and actual retention or churn as target. We obtained the following two discrimination rules: 'IF ANN_I output value <0.285, THEN Final Solution = Retention' and 'IF ANN_I output value ${\geq}0.285$, THEN Final Solution = Churn.' The threshold 0.285 is the value optimized for the data used in this research. The result we present in this research is the structure or framework of our hybrid SVM model, not a specific threshold value such as 0.285. Therefore, the threshold value in the above discrimination rules can be changed to any value depending on the data. In order to evaluate the performance of our hybrid SVM model, we used the 'churn data set' in UCI Machine Learning Repository, that consists of 85% retention customers and 15% churn customers. Accuracy of the hybrid SVM model is 91.08% that is better than that of SVM_I model or SVM_B model. The points worth noticing here are its sensitivity, 95.02%, and specificity, 69.24%. The sensitivity of SVM_I model is 94.65%, and the specificity of SVM_B model is 67.00%. Therefore the hybrid SVM model developed in this research improves the specificity of SVM_B model while maintaining the sensitivity of SVM_I model.