• Journal of Intelligence and Information Systems
• /
• v.23 no.1
• /
• pp.95-108
• /
• 2017

Recently, AlphaGo which is Bakuk (Go) artificial intelligence program by Google DeepMind, had a huge victory against Lee Sedol. Many people thought that machines would not be able to win a man in Go games because the number of paths to make a one move is more than the number of atoms in the universe unlike chess, but the result was the opposite to what people predicted. After the match, artificial intelligence technology was focused as a core technology of the fourth industrial revolution and attracted attentions from various application domains. Especially, deep learning technique have been attracted as a core artificial intelligence technology used in the AlphaGo algorithm. The deep learning technique is already being applied to many problems. Especially, it shows good performance in image recognition field. In addition, it shows good performance in high dimensional data area such as voice, image and natural language, which was difficult to get good performance using existing machine learning techniques. However, in contrast, it is difficult to find deep leaning researches on traditional business data and structured data analysis. In this study, we tried to find out whether the deep learning techniques have been studied so far can be used not only for the recognition of high dimensional data but also for the binary classification problem of traditional business data analysis such as customer churn analysis, marketing response prediction, and default prediction. And we compare the performance of the deep learning techniques with that of traditional artificial neural network models. The experimental data in the paper is the telemarketing response data of a bank in Portugal. It has input variables such as age, occupation, loan status, and the number of previous telemarketing and has a binary target variable that records whether the customer intends to open an account or not. In this study, to evaluate the possibility of utilization of deep learning algorithms and techniques in binary classification problem, we compared the performance of various models using CNN, LSTM algorithm and dropout, which are widely used algorithms and techniques in deep learning, with that of MLP models which is a traditional artificial neural network model. However, since all the network design alternatives can not be tested due to the nature of the artificial neural network, the experiment was conducted based on restricted settings on the number of hidden layers, the number of neurons in the hidden layer, the number of output data (filters), and the application conditions of the dropout technique. The F1 Score was used to evaluate the performance of models to show how well the models work to classify the interesting class instead of the overall accuracy. The detail methods for applying each deep learning technique in the experiment is as follows. The CNN algorithm is a method that reads adjacent values from a specific value and recognizes the features, but it does not matter how close the distance of each business data field is because each field is usually independent. In this experiment, we set the filter size of the CNN algorithm as the number of fields to learn the whole characteristics of the data at once, and added a hidden layer to make decision based on the additional features. For the model having two LSTM layers, the input direction of the second layer is put in reversed position with first layer in order to reduce the influence from the position of each field. In the case of the dropout technique, we set the neurons to disappear with a probability of 0.5 for each hidden layer. The experimental results show that the predicted model with the highest F1 score was the CNN model using the dropout technique, and the next best model was the MLP model with two hidden layers using the dropout technique. In this study, we were able to get some findings as the experiment had proceeded. First, models using dropout techniques have a slightly more conservative prediction than those without dropout techniques, and it generally shows better performance in classification. Second, CNN models show better classification performance than MLP models. This is interesting because it has shown good performance in binary classification problems which it rarely have been applied to, as well as in the fields where it's effectiveness has been proven. Third, the LSTM algorithm seems to be unsuitable for binary classification problems because the training time is too long compared to the performance improvement. From these results, we can confirm that some of the deep learning algorithms can be applied to solve business binary classification problems.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.105-129
• /
• 2020

This study uses corporate data from 2012 to 2018 when K-IFRS was applied in earnest to predict default risks. The data used in the analysis totaled 10,545 rows, consisting of 160 columns including 38 in the statement of financial position, 26 in the statement of comprehensive income, 11 in the statement of cash flows, and 76 in the index of financial ratios. Unlike most previous prior studies used the default event as the basis for learning about default risk, this study calculated default risk using the market capitalization and stock price volatility of each company based on the Merton model. Through this, it was able to solve the problem of data imbalance due to the scarcity of default events, which had been pointed out as the limitation of the existing methodology, and the problem of reflecting the difference in default risk that exists within ordinary companies. Because learning was conducted only by using corporate information available to unlisted companies, default risks of unlisted companies without stock price information can be appropriately derived. Through this, it can provide stable default risk assessment services to unlisted companies that are difficult to determine proper default risk with traditional credit rating models such as small and medium-sized companies and startups. Although there has been an active study of predicting corporate default risks using machine learning recently, model bias issues exist because most studies are making predictions based on a single model. Stable and reliable valuation methodology is required for the calculation of default risk, given that the entity's default risk information is very widely utilized in the market and the sensitivity to the difference in default risk is high. Also, Strict standards are also required for methods of calculation. The credit rating method stipulated by the Financial Services Commission in the Financial Investment Regulations calls for the preparation of evaluation methods, including verification of the adequacy of evaluation methods, in consideration of past statistical data and experiences on credit ratings and changes in future market conditions. This study allowed the reduction of individual models' bias by utilizing stacking ensemble techniques that synthesize various machine learning models. This allows us to capture complex nonlinear relationships between default risk and various corporate information and maximize the advantages of machine learning-based default risk prediction models that take less time to calculate. To calculate forecasts by sub model to be used as input data for the Stacking Ensemble model, training data were divided into seven pieces, and sub-models were trained in a divided set to produce forecasts. To compare the predictive power of the Stacking Ensemble model, Random Forest, MLP, and CNN models were trained with full training data, then the predictive power of each model was verified on the test set. The analysis showed that the Stacking Ensemble model exceeded the predictive power of the Random Forest model, which had the best performance on a single model. Next, to check for statistically significant differences between the Stacking Ensemble model and the forecasts for each individual model, the Pair between the Stacking Ensemble model and each individual model was constructed. Because the results of the Shapiro-wilk normality test also showed that all Pair did not follow normality, Using the nonparametric method wilcoxon rank sum test, we checked whether the two model forecasts that make up the Pair showed statistically significant differences. The analysis showed that the forecasts of the Staging Ensemble model showed statistically significant differences from those of the MLP model and CNN model. In addition, this study can provide a methodology that allows existing credit rating agencies to apply machine learning-based bankruptcy risk prediction methodologies, given that traditional credit rating models can also be reflected as sub-models to calculate the final default probability. Also, the Stacking Ensemble techniques proposed in this study can help design to meet the requirements of the Financial Investment Business Regulations through the combination of various sub-models. We hope that this research will be used as a resource to increase practical use by overcoming and improving the limitations of existing machine learning-based models.

• Journal of Intelligence and Information Systems
• /
• v.16 no.2
• /
• pp.19-32
• /
• 2010

Volatility plays a central role in both academic and practical applications, especially in pricing financial derivative products and trading volatility strategies. This study presents a novel mechanism based on generalized autoregressive conditional heteroskedasticity (GARCH) models that is able to enhance the performance of intelligent volatility trading systems by predicting Korean stock market volatility more accurately. In particular, we embedded the concept of the volatility asymmetry documented widely in the literature into our model. The newly developed Korean stock market volatility index of KOSPI 200, VKOSPI, is used as a volatility proxy. It is the price of a linear portfolio of the KOSPI 200 index options and measures the effect of the expectations of dealers and option traders on stock market volatility for 30 calendar days. The KOSPI 200 index options market started in 1997 and has become the most actively traded market in the world. Its trading volume is more than 10 million contracts a day and records the highest of all the stock index option markets. Therefore, analyzing the VKOSPI has great importance in understanding volatility inherent in option prices and can afford some trading ideas for futures and option dealers. Use of the VKOSPI as volatility proxy avoids statistical estimation problems associated with other measures of volatility since the VKOSPI is model-free expected volatility of market participants calculated directly from the transacted option prices. This study estimates the symmetric and asymmetric GARCH models for the KOSPI 200 index from January 2003 to December 2006 by the maximum likelihood procedure. Asymmetric GARCH models include GJR-GARCH model of Glosten, Jagannathan and Runke, exponential GARCH model of Nelson and power autoregressive conditional heteroskedasticity (ARCH) of Ding, Granger and Engle. Symmetric GARCH model indicates basic GARCH (1, 1). Tomorrow's forecasted value and change direction of stock market volatility are obtained by recursive GARCH specifications from January 2007 to December 2009 and are compared with the VKOSPI. Empirical results indicate that negative unanticipated returns increase volatility more than positive return shocks of equal magnitude decrease volatility, indicating the existence of volatility asymmetry in the Korean stock market. The point value and change direction of tomorrow VKOSPI are estimated and forecasted by GARCH models. Volatility trading system is developed using the forecasted change direction of the VKOSPI, that is, if tomorrow VKOSPI is expected to rise, a long straddle or strangle position is established. A short straddle or strangle position is taken if VKOSPI is expected to fall tomorrow. Total profit is calculated as the cumulative sum of the VKOSPI percentage change. If forecasted direction is correct, the absolute value of the VKOSPI percentage changes is added to trading profit. It is subtracted from the trading profit if forecasted direction is not correct. For the in-sample period, the power ARCH model best fits in a statistical metric, Mean Squared Prediction Error (MSPE), and the exponential GARCH model shows the highest Mean Correct Prediction (MCP). The power ARCH model best fits also for the out-of-sample period and provides the highest probability for the VKOSPI change direction tomorrow. Generally, the power ARCH model shows the best fit for the VKOSPI. All the GARCH models provide trading profits for volatility trading system and the exponential GARCH model shows the best performance, annual profit of 197.56%, during the in-sample period. The GARCH models present trading profits during the out-of-sample period except for the exponential GARCH model. During the out-of-sample period, the power ARCH model shows the largest annual trading profit of 38%. The volatility clustering and asymmetry found in this research are the reflection of volatility non-linearity. This further suggests that combining the asymmetric GARCH models and artificial neural networks can significantly enhance the performance of the suggested volatility trading system, since artificial neural networks have been shown to effectively model nonlinear relationships.

• Journal of Korean Society of Forest Science
• /
• v.75 no.1
• /
• pp.32-37
• /
• 1986

In order to get the fundamental information that could be useful to judge the potentiality of occurrence of rapid shallow landslide in the objective slope, factors selected on Jinhae regions in Korea, where many landslides were caused by heavy rainfall of daily 465 mm and hourly 52mm in August 1979, was carried out through the multiple statistics of quantification method (II) by the electronic computer. The net system with $2{\times}2cm$ unit mesh was overlayed with the contour map of scale 1:5000. 74 meshes of landslides and 119 meshes of non-landslide were sampled out to survey the state of vegetative cover and geomorphological conditions, those were divided into 6 items arid 27 categories. As a result, main factors that would lead to landslide were shown in order of vegetation, slope type, slope position, slope, aspect and numbers of stream. Particularly, coniferous forest of 10 years old, concave slope and foot of mountain were main factors making slope instability. On the contrary, coniferous forest of 20-30 years old, deciduous forest, convex slope and summit contributed to the stable against Landslide. The boundary value between two groups of existence and none of landslides was -0.123, and its prediction was 72%. It was well predicted to divide into two groups of them.

• Journal of the Ergonomics Society of Korea
• /
• v.36 no.1
• /
• pp.23-36
• /
• 2017

Objective: This paper presents a study to evaluate the WBGT index for assessing the effects of a wide range of outdoor weather conditions on human responses. Background: The Wet Bulb Globe Temperature (WBGT) index was firstly developed for the assessment of hot outdoor conditions. It is a recognised index that is used world-wide. It may be useful over a range of outdoor conditions and not just for hot climates. Method: Four group experiments, involving people performing a light stepping activity, were conducted to determine human responses to outside conditions in the U.K. They were conducted in September 2007 (autumn), December 2007 (winter), March 2008 (spring) and June 2008 (summer). Environmental measurements included WBGT, air temperature, radiant temperature (including solar load), humidity and wind speed all measured at 1.2m above the ground, as well as weather data measured by a standard weather station at 3m to 4m above the ground. Participants' physiological and subjective responses were measured. When the overall results of the four seasons are considered, WBGT provided a strong prediction of physiological responses as well as subjective responses if aural temperature, heart rate and sweat production were measured. Results: WBGT is appropriate to predict thermal strain on a large group of ordinary people in moderate conditions. Consideration should be given to include the WBGT index in warning systems for a wide range of weather conditions. However, the WBGT overestimated physiological responses of subjects. In addition, tenfold Borg's RPE was significantly different with heart rate measured for the four conditions except autumn (p<0.05). Physiological and subjective responses over 60 minutes consistently showed a similar tendency in the relationships with the $WBGT_{head}$ and $WBGT_{abdomen}$. Conclusion: It was found that either $WBGT_{head}$ or $WBGT_{abdomen}$ could be measured if a measurement should be conducted at only one height. The relationship between the WBGT values and weather station data was also investigated. There was a significant relationship between WBGT values at the position of a person and weather station data. For UK daytime weather conditions ranging from an average air temperature of $6^{\circ}C$ to $21^{\circ}C$ with mean radiant temperatures of up to $57^{\circ}C$, the WBGT index could be used as a simple thermal index to indicate the effects of weather on people. Application: The result of evaluation of WBGT might help to develop the smart clothing for workers in industrial sites and improve the work environment in terms of considering workers' wellness.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• v.25 no.4
• /
• pp.324-329
• /
• 1999

The facial esthetics are much affected by nasal changes due to especially its central position in relation to facial outline and so appropriately evaluated should be the functional and esthetic aspects of the nose associated with the facial appearance. Generally, a maxillary surgical movement is known to induce the changes of nasolabial morphology secondary to the skeletal repositioning accompanied by muscular retraction. These changes can be desirable or undesirable to individuals according to the direction and amount of maxillary repositioning. We investigated the surgical changes of bony maxilla and its effects to nasal morphology through the analysis of the lateral cephalogram in the Le Fort I osteotomy. Subjects were 10 patients(male 2, female 8, mean age 22.3 years) and cephalograms were obtained 2 weeks before surgery(T1) and 6 months after surgery(T2). The surgical maxillary movement was identified through the horizontal and vertical repositioning of point A. Soft-tissue analysis of the nasal profile was performed employing two angles: nasal tip projection(NTP), columellar angle(CA). Also, alar base width(ABW) was assessed directly on the patients with a slide gauge. The results were as follows; 1. Both anterior and superior movement above 2mm of maxilla rotated up nasal tip above 1mm. Either anterior or superior movement above 2mm of maxilla made prediction of the amount & direction of NTP changes difficult. Especially, a correlation between horizontal movement of maxilla and NTP rotated-up was P<0.01. 2. Both much highly anterior and superior movement of maxilla is accompanied by more CA increase than either highly. Especially, the correlation between horizontal movement of maxilla and CA change was P<0.05. 3. Anterior and/or superior movement of maxilla was accompanied by the unpredictable ABW widening. 4. The amount of changes of NTP, CA, and ABW is not in direct proportion to amout of anterior and/or superior movement of maxilla. 5. Nasal morphologic changes following Le Fort I osteotomy are affacted by not merely bony repositioning but other multiple factors.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.18 no.2
• /
• pp.57-61
• /
• 2014

Purpose Iodine (I-131) is one of the most widely used radioactive isotopes for therapeutic in the field of nuclear medicine. Therapeutic I-131 capsule is made out of lead to shield high energy radiation. Accurate dosimetry is necessarily required to perform safe and effective work for relative workers. The Monte Carlo method is known as a method to predict the absorbed dose distribution most accurately in radiation therapy and many researchers constantly attempt to apply this method to the dose calculation of radiotherapy recently. This paper aims to calculate distance dependent and activity dependent therapeutic I-131 capsule using GEANT4. Materials and Methods Therapeutic capsules was implemented on the basis of the design drawings. The simulated dose was determined by generating of gamma rays of energy to more than 364 keV. The simulated dose from the capsule at the distance of 10 cm and 100 cm was measured and calculated in the model of water phantom. The simulated dose were separately calculated for each position of each detector. Results According to the domestic regulation on radiation safety, the dose at 10 cm and 100 cm away from the surface of therapeutic I-131 capsule should not exceed 2.0 mSv/h and 0.02 mSv/h, respectively. The simulated doses turned out to be less than the limit, satisfying the domestic regulation. Conclusion These simulation results may serve as useful data in the prediction of hands dose absorbed by I-131 capsule handling. GEANT4 is considered that it will be effectively used in order to check the radiation dose.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.1061-1061
• /
• 2001

The scope of precision agriculture is to reach the put up cultivation goals by adjusting inputs as precise as possible after what is required by the soil and crop potentials, on a high spatial resolution. Consequently, precision agriculture is also often called site specific agriculture. Regulation of field inputs “on the run” has been made possible by the GPS (Geographical Position System)-technology, which gives the farmer his exact real time positioning in the field. The general goal with precision agriculture is to apply inputs where they best fill their purpose. Thus, resources could be saved, and nutrient losses as well as the impact on the environment could be minimized without lowering total yields or putting product quality at risk. As already indicated the technology exists to regulate the input based on beforehand decisions. However, the real challenge is to provide a reliable basis for decision-making. To support high spatial resolution, extensive sampling and analysis is required for many soil and plant characteristics. The potential of the NIR-technology to provide rapid, low cost analyses with a minimum of sample preparation for a multitude of characteristics therefore constitutes a far to irresistible opportunity to be un-scrutinized. In our work we have concentrated on soil-analysis. The instrument we have used is a Bran Lubbe InfraAlyzer 500 (1300-2500 nm). Clay- and organic matter-contents are soil constituents with major implications for most properties and processes in the soil system. For these constituents we had a 3000-sample material provided. High performance models for the agricultural areas in Sweden have been constructed for clay-content, but a rather large reference material is required, probably due to the large variability of Swedish soils. By subdividing Sweden into six areas the total performance was improved. Unfortunately organic matter was not as easy to get at. Reliable models for larger areas could not be constructed. However, through keeping the mineral fraction of the soil at minimal variation good performance could be achieved locally. The influence of a highly variable mineral fraction is probably one of the reasons for the contradictory results found in the literature regarding organic matter content. Tentative studies have also been performed to elucidate the potential performance in contexts with direct operational implications: lime requirement and prediction of plant uptake of soil nitrogen. In both cases there is no definite reference method, but there are numerous indirect, or indicator, methods suggested. In our study, field experiments where used as references and NIR was compared with methods normally used in Sweden. The NIR-models performed equally or slightly better as the standard methods in both situations. However, whether this is good enough is open for evaluation.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.4 no.2
• /
• pp.74-83
• /
• 2001

The purpose of this study was investigate to the influence of forest roads characteristics and environment factors on the soil erosion, stability and vegetation survival of cut slope in forest roads. The results obtained could be summarized as follows; 1. The correlated factors between slope erosion and variables in cut slope were altitude, convex, degree of slope, length of slope and soil depth. In the stepwise regression analysis, length of slope and soil hardness was a high significant and its regression equation was given by -89.6136 + 15.0667X14 + 16.6713X15($R^2$ = 0.6712). 2. The main factors influencing the stability of cut slope were significant in order of coverage, middle, convex, length of slope and north, and its discriminant equation was given by -1.019 + 0.064X22 - 0.808X8 - 0.622X24 + 0.742X11 - 0.172X14 - 0.545X6 ($R^2$ = 0.793). 3. The centroids value of discriminant function in the stability and unstability estimated to 1.244 and -1.348, respectively. The boundary value between two groups related to slope stability was -0.1038. The prediction rate of discriminant function for stability evaluation of was as high as 91.3%. 4. The dominant species of invasion vegetation on the cut slope consist with Carex humilis, Agropyron tsukushiense var. transiens, Calamagrostis arundinacea, Miscanthus sinensis var. purpurascens, and Ixeris dentata in survey area. The rate of vegetation invasion more increased by time passed. 5. The life form of invasion vegetation in cut slop showed to $H-D_1-R_{2,3}-e$ type of the hemicryptophyte of dormancy form, dissem inated widely by wind and water of dissminule type, moderate extent and narrowest extent of radicoid type, erect form of growth form. 6. The correlated factors between forest enviroment and coverage appeared north, passage years and middle position of slope at 5% level. The forest environment factors influencing the invasion plants in survey area were shown in order to altitude, passage years, rock(none), forest type(mixed) and stone amount. The regression equation was given by 17.5228 - 0.0911X3 + 3.6189X28 15.8493X22 19.8544X25 + 0.3558X26 ($R^2$ = 0.4026).

• Journal of Astronomy and Space Sciences
• /
• v.22 no.3
• /
• pp.263-272
• /
• 2005

For the Communication, Ocean and Meteorological Satellite (COMS) which will be launched in 2008, an algorithm for finding the precise location of the sun-glint point on the ocean surface is studied. The precise locations of the sun-glint are estimated by considering azimuth and elevation angles of Sun-satellite-Earth geometric position and the law of reflection. The obtained nonlinear equations are solved by using the Newton-Raphson method. As a result, when COMS is located at $116.2^{\circ}E$ or $128.2^{\circ}E$ longitude, the sun-glint covers region of ${\pm}10^{\circ}(N-S)$ latitude and $80-150^{\circ}(E-W)$ longitude. The diurnal path of the sun-glint in the southern hemisphere is curved towards the North Pole, and the path in the northern hemisphere is forwards the south pole. The algorithm presented in this paper can be applied to predict the precise location of sun-glint region in any other geostationary satellites.