• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.173-198
• /
• 2020

For a long time, many studies have been conducted on predicting the success of campaigns for customers in academia, and prediction models applying various techniques are still being studied. Recently, as campaign channels have been expanded in various ways due to the rapid revitalization of online, various types of campaigns are being carried out by companies at a level that cannot be compared to the past. However, customers tend to perceive it as spam as the fatigue of campaigns due to duplicate exposure increases. Also, from a corporate standpoint, there is a problem that the effectiveness of the campaign itself is decreasing, such as increasing the cost of investing in the campaign, which leads to the low actual campaign success rate. Accordingly, various studies are ongoing to improve the effectiveness of the campaign in practice. This campaign system has the ultimate purpose to increase the success rate of various campaigns by collecting and analyzing various data related to customers and using them for campaigns. In particular, recent attempts to make various predictions related to the response of campaigns using machine learning have been made. It is very important to select appropriate features due to the various features of campaign data. If all of the input data are used in the process of classifying a large amount of data, it takes a lot of learning time as the classification class expands, so the minimum input data set must be extracted and used from the entire data. In addition, when a trained model is generated by using too many features, prediction accuracy may be degraded due to overfitting or correlation between features. Therefore, in order to improve accuracy, a feature selection technique that removes features close to noise should be applied, and feature selection is a necessary process in order to analyze a high-dimensional data set. Among the greedy algorithms, SFS (Sequential Forward Selection), SBS (Sequential Backward Selection), SFFS (Sequential Floating Forward Selection), etc. are widely used as traditional feature selection techniques. It is also true that if there are many risks and many features, there is a limitation in that the performance for classification prediction is poor and it takes a lot of learning time. Therefore, in this study, we propose an improved feature selection algorithm to enhance the effectiveness of the existing campaign. The purpose of this study is to improve the existing SFFS sequential method in the process of searching for feature subsets that are the basis for improving machine learning model performance using statistical characteristics of the data to be processed in the campaign system. Through this, features that have a lot of influence on performance are first derived, features that have a negative effect are removed, and then the sequential method is applied to increase the efficiency for search performance and to apply an improved algorithm to enable generalized prediction. Through this, it was confirmed that the proposed model showed better search and prediction performance than the traditional greed algorithm. Compared with the original data set, greed algorithm, genetic algorithm (GA), and recursive feature elimination (RFE), the campaign success prediction was higher. In addition, when performing campaign success prediction, the improved feature selection algorithm was found to be helpful in analyzing and interpreting the prediction results by providing the importance of the derived features. This is important features such as age, customer rating, and sales, which were previously known statistically. Unlike the previous campaign planners, features such as the combined product name, average 3-month data consumption rate, and the last 3-month wireless data usage were unexpectedly selected as important features for the campaign response, which they rarely used to select campaign targets. It was confirmed that base attributes can also be very important features depending on the type of campaign. Through this, it is possible to analyze and understand the important characteristics of each campaign type.

• Journal of Intelligence and Information Systems
• /
• v.23 no.4
• /
• pp.147-168
• /
• 2017

There have been many studies on accurate stock market forecasting in academia for a long time, and now there are also various forecasting models using various techniques. Recently, many attempts have been made to predict the stock index using various machine learning methods including Deep Learning. Although the fundamental analysis and the technical analysis method are used for the analysis of the traditional stock investment transaction, the technical analysis method is more useful for the application of the short-term transaction prediction or statistical and mathematical techniques. Most of the studies that have been conducted using these technical indicators have studied the model of predicting stock prices by binary classification - rising or falling - of stock market fluctuations in the future market (usually next trading day). However, it is also true that this binary classification has many unfavorable aspects in predicting trends, identifying trading signals, or signaling portfolio rebalancing. In this study, we try to predict the stock index by expanding the stock index trend (upward trend, boxed, downward trend) to the multiple classification system in the existing binary index method. In order to solve this multi-classification problem, a technique such as Multinomial Logistic Regression Analysis (MLOGIT), Multiple Discriminant Analysis (MDA) or Artificial Neural Networks (ANN) we propose an optimization model using Genetic Algorithm as a wrapper for improving the performance of this model using Multi-classification Support Vector Machines (MSVM), which has proved to be superior in prediction performance. In particular, the proposed model named GA-MSVM is designed to maximize model performance by optimizing not only the kernel function parameters of MSVM, but also the optimal selection of input variables (feature selection) as well as instance selection. In order to verify the performance of the proposed model, we applied the proposed method to the real data. The results show that the proposed method is more effective than the conventional multivariate SVM, which has been known to show the best prediction performance up to now, as well as existing artificial intelligence / data mining techniques such as MDA, MLOGIT, CBR, and it is confirmed that the prediction performance is better than this. Especially, it has been confirmed that the 'instance selection' plays a very important role in predicting the stock index trend, and it is confirmed that the improvement effect of the model is more important than other factors. To verify the usefulness of GA-MSVM, we applied it to Korea's real KOSPI200 stock index trend forecast. Our research is primarily aimed at predicting trend segments to capture signal acquisition or short-term trend transition points. The experimental data set includes technical indicators such as the price and volatility index (2004 ~ 2017) and macroeconomic data (interest rate, exchange rate, S&P 500, etc.) of KOSPI200 stock index in Korea. Using a variety of statistical methods including one-way ANOVA and stepwise MDA, 15 indicators were selected as candidate independent variables. The dependent variable, trend classification, was classified into three states: 1 (upward trend), 0 (boxed), and -1 (downward trend). 70% of the total data for each class was used for training and the remaining 30% was used for verifying. To verify the performance of the proposed model, several comparative model experiments such as MDA, MLOGIT, CBR, ANN and MSVM were conducted. MSVM has adopted the One-Against-One (OAO) approach, which is known as the most accurate approach among the various MSVM approaches. Although there are some limitations, the final experimental results demonstrate that the proposed model, GA-MSVM, performs at a significantly higher level than all comparative models.

• Journal of Chest Surgery
• /
• v.35 no.3
• /
• pp.188-198
• /
• 2002

Background: Reoperations on the aortic root or the ascending aorta are being performed with increasing frequency and remain a challenging problem. This study was performed to analyze the results of reoperations on the ascending aorta and aortic root. Material and Method: Between May 1995 and April 2001, 30 patients had reoperations on the ascending aorta and aortic root and were reviewed retrospectively. The mean interval between the previous repair and the actual reoperation was 56 months(range 3 to 142 months). Seven patients(23.3%) had two or more previous operations. The indications for reoperations were true aneurysm in 7 patients(23.3%), prosthetic valve endocarditis in 6(20%), false aneurysm in 5(16.7%), paravalvular leak associated with Behcet's disease in 4(13.3%), malfunction of prosthetic aortic valve in 4(13.3%), aortic dissection in 3(10%), and annuloaortic ectasia in 1(3.3%). The principal reoperations performed were aortic root replacement in 17 patients(56.7%), replacement of the ascending aorta in 8(26.7%), aortic and mitral valve replacement with reconstruction of fibrous trigone in 2(6.6%), patch aortoplasty in 2(6.6%), and aortic valve replacement after Bentall operation in 1 (3.3%). The cardiopulmonary bypass was started before sternotomy in 7 patients and the hypothermic circulatory arrest was used in 16(53.3%). The mean time of circulatory arrest, total bypass, and aortic crossclamp were 20$\pm$ 12 minutes, 228$\pm$56 minutes, and 143$\pm$62 minutes, respectively Result: There were three early deaths(10%). The postoperative complications were reoperation for bleeding in 7 patients(23.3%), cardiac complications in 5(16.7%), transient acute renal failure in 2(6.6%), transient focal seizure in 2(6.6%), and the others in 5. The mean follow-up was 22.8 $\pm$20.5 months. There were two late deaths(7.4%). The actuarial survival was 92.6$\pm$5.0% at 6 years. One patient required reoperation for complication of reoperation on the ascending aorta and aortic root(3.7%). The 1- and 6-year actuarial freedom from reoperation was 100% and 83.3$\pm$15.2%, respectively. One patient with Behcet's disease are waiting for reoperation due to false aneurysm, which developed after aortic root replacement with homograft. There were no thromboembolisms or anticoagulant related complications. Conclusions: This study suggests that reoperations on the ascending aorta and aortic root can be performed with acceptable early mortality and morbidity, and adequate surgical strategies according to the pathologi conditions are critical to the prevention of the reoperation.

• Korean journal of applied entomology
• /
• v.15 no.3 s.28
• /
• pp.147-151
• /
• 1976

Laboratory experiments were conducted to evaluate the resistance of the new rice varieties Milyang $\#$21 and $\#$23 at the seedling stage to the brown planthopper(Nilaparvata lugens), small brown planthopper(Laodelphax striatellus), white-back planthopper(Sogatella furcifera), green rorice leafhopper(Nephotettix cincticeps) and zigzag-striped leafhopper (Recilia dorsalis) The varieties Yushin and T(N)-1 were used as susceptible check and the resistant check varieties were Mudgo to brown planthopper, ASD-7 :o small brown planthopper, Colombo to white-back planthopper, IR2061 (46763) to green rice leafhopper and Vellailanalgayan to zigzag-striped leafhopper. The varieties Milyang $\#$21 and $\#$23 were moderate in plant reaction only to the green rice leafhopper and $\#$23 was moderately resistant in plant rection only to the small brown planthopper. The nymphs of plant-and leaf-hoppers were more prefered Milyang $\#$21 and $\#$23 and susceptible check-varieties were more preferred for feeding than the resistant check-varieties. The green rice leafhopper and aigzag·striped leafhopper much more prefered Milyang $\#$21 and $\#$23 for oviposition, while brown planthopper and small brown planthopper more prefered the resistant cheeks than test varieties and susceptible checks. However, there was no any steady relationship in ovipositional preference between resistant and susceptible to the hoppers. Shorter nymphal-periods and higher rate of adult emergence were observed in the test varieties and the susceptible checks compared with the resistant checks. In conclusion, the varieties Milyang $\#$21 and $\#$23 seemed to be lack of true resistance to the plant-and leaf-hoppers from the viewpoints of antibiosis and feeding preference.

• Journal of Environmental Health Sciences
• /
• v.43 no.1
• /
• pp.1-22
• /
• 2017

'The worstest environment disaster', 'World's first biocide massacre', 'Home-based Sewol ferry disaster' are all phrases attached to the recent humidifier disinfectant disaster. In the spring of 2011, four of 8 pregnant women including 1 adult man passed away at a university hospital in Seoul due to breathing failure. Epidemiologic investigation conducted by the Korean CDC soon revealed the inhalation of humidifier disinfectant, which had been widely used in Korea during the winter, to be responsible for the disease. As well as lung fibrosis hardening of the lungs, other diseases including asthma, rhinitis, skin disease, liver disease, fetal disease or cancers have been researched for their relation with exposure to the products. By February 9, 2017, 5,342 cases had registered for health problems and 1,131 of them were already dead (20.8% mortality rate). Based on studies by government agencies and a telephone survey of the general population by Seoul National University and civic groups, around 20% of the general public of Korea has used these products. Since the market release of the first product by SK Chemical in 1994, over 7.1 million items from around 20 brands were sold up to 2011. Most of the products were manufactured by well-known large conglomerates such as SK, Lotte, Samsung, Shinsegye, LG, and GS, as well as some European companies including UK-based Reckitt Benckiser and TESCO, the German firm Henkel, the Danish firm KeTox, and an Irish company. Even though this disaster was unveiled in 2011 by the Korean government, the issue of the victims was neglected for over five years. In 2016, an unexpected but intensive investigation by prosecutors found that Reckitt Benckiser manipulated and concealed animal tests for its own brand and brought several university experts and company employees to court. The matter was an intense social issue in Korea from May to June with a surge in media coverage. The prosecutor's investigation and a nationwide boycott campaign organized by victims and environmental groups against Reckitt Benckiser, whose product had been used by more than 70% of victims, led to the producer's official apology and a compensation scheme. A legislative investigation organized after the April 2016 national election revealed the producers' faults and the government's responsibility, but failed to meet expectations. A special law for the victims passed the National Assembly in January 2017 and a punitive system together with a massive environmental epidemiology investigation are expected to be the only solutions for this tragedy. Sciences of medicine, toxicology and environmental health have provided decisive evidence so far, but for the remaining problems the perspectives of social sciences such as sociology and jurisprudence are highly necessary, similar to with the Minamata disease and Wonjin Rayon events. It may not be easy to follow this issue using unfamiliar terminology from medical and chemical science and the long, complicated history of the event. For these reasons the author has attempted to write this article in a question and answer format to render it easier to follow. The 17 questions are: Q1 What is humidifier disinfectant? Q2 What kind of health problems are caused by humidifier disinfectant? Q3 How many victims are there? Q4 What is the analysis of the 1,112 cases of death? Q5 What is the problem with the government's diagnostic criteria and the solution? Q6 Who made what brands? Q7 Has there been a recall? What is still on sale? Q8 Was safety not checked by any producers? Q9 What are the government's responsibilities? Q10 Is it true that these products were sold only in Korea? Q11 Why and how was it unveiled only in 2011 after 17 years of sales? Q12 What delayed the resolution of the victim issue? Q13 What is the background of the prosecutor's investigation in early 2016? Q14 Is it possible to report new victim cases without evidence of product purchase? Q15 What is happening with the victim issue? Q16 How does it compare with the cases of Minamata disease and Wonjin Rayon? Q17 Are there prevention measures and lessons?

• Korean Journal of Environmental Biology
• /
• v.22 no.2
• /
• pp.247-258
• /
• 2004

Climatological, hydrological and planktonical research studies, measurements of primary production and photosynthetic efficiency from December 1976 to December 1978 have been carried out in two brackish lakes: Lake Etang de Berre and Lake Etang de Vaine located in the French Mediterranean coast, in the region of Carry-le-Rouet located on the north-west Mediterranean near Marseilles, and in fresh water inflows from 4 Rivers (Touloubre, Durance, Arc, Durancole) to Lake Etang de Berre. Physico-chemical parameters were measured for this study: water temperature, salinity, density, pH, alcalinity, dissolved oxygen (% saturation), phosphate, nitrate, nitrite, silicate etc. Diverse biological parameters were also studied: photosynthetic pigments, phaeopigments, specific composition and biomass of phytoplankton, primary pelagic production etc. Climatical factors were studied: air-temperature, solar-radiation, evaporation, direction (including strength) of winds, precipitation and freshwater volume of the four rivers. The changes in Lake ‘Etang de Berre’ ecosystem depend on the quality of the water in the Durance River, and on the effects of seawater near the entrance of the Caronte Canal. The water quality of the lake varies horizontally and vertically as a result of atmospheric phenomena, maritime currents and tides. The distribution of water temperatures is generally heterogeneous. Southeasterly winds and the Northeasterly Mistral wind are important in the origins of circulated and mixed water masses. These winds are both frequent and strong. They have, as a result, a great effect on the water environment of Lake Etang de Berre. In theory, the annual precipitation in this region is well over eight times the water mass of the lake. The water of the Durance River flows into Lake Etang de Berre through the EDF Canal, amounting to 90% of the precipitation. However, reduction of rainfall in dry seasons has a serious effect on the hydrological characteristics of the lake. The temperature in the winter is partially caused by the low temperature of fresh water, particularly that of the Durance River. The hydrological season of fresh and brackish water is about one month ahead of the hydrological season of sea water in its vicinity. The salinity of Lake Etang de Berre runs approximately 3$\textperthousand$, except at lower levels and near the entrance to the Caronte Canal. However, when the volume of the Durance River water is reduced in the summer and fall, the salinity rises to 15$\textperthousand$. In the lake, the ratio of fresh water to sea water is six to one (6:1). The large quantities of seston conveyed by rivers, particularly the Durance diversion, strongly reduce the transparency in the brackish waters. Although the amount of sunshine is also notable, transparency is slight because of the large amount of seston, carried chiefly by Tripton in the fresh water of the Durance River. Therefore, photosynthesis generally occurs only in the surface layer. The transparency progressively increases from freshwater to open seawater, as mineral particles sink to the bottom (about 1.7kg $m^{-2}a^{-1}$ on the average in brackish lakes). The concentration of dissolved oxygen and the rate of oxygen saturation in seawater (Carry-le-Rouet) ranged from 5.0 to 6.0 $m\ell$ㆍ.$1^{-1}$, and from 95 to 105%, respectively. The amount of dissolved oxygen in Etang de Berre oscillated between 2.9 and 268.3%. The monographs of phosphate, nitrate, nitrite and silicate were published as a part of a study on the ecology of phytoplankton in these environments. Horizontal and vertical distributions of these nutriments were studied in detail. The recent diversion of the Durance River into Lake Etang de Berre has effected a fundamental change in this formerly marine environment, which has had a great impact in its plankton populations. A total of 182 taxa were identified, including 111 Bacillariophyceae, 44 Chlorophyceae, and 15 Cyanophyceae. The most abundant species are small freshwater algae, mainly Chlorophyceae. The average density is about $10^{8}$ cells $1^{-1}$ in Lake Etang de Berre, and about double that amount in Lake Etang de Vaine. Differences in phytoplankton abundance and composition at the various stations or at various depths are slight. Cell biovolume V (equivalent to true biomass), plasma volume VP (‘useful’ biomass) and, simultaneously. the cell surface area S and S/V ratio through the measurement of cell dimensions were computed as the parameters of phytoplankton productivity and metabolism. Pigment concentrations are generally very high on account of phytoplankton blooms by Cyanophyceae, Chlorophyceae and Cryptophyceae. On the other hand, in freshwaters and marine waters, pigment concentrations are comparatively low and stable, showing slight annual variation. The variations of ATP concentration were closely related to those of chlorophyll a and phytoplankton blooms only in marine waters. The carbon uptake rates ranged between 38 and 1091 mg$Cm^{-2}d^{-1}$, with an average surface value of 256 mg; water-column carbon-uptake rates ranged between 240 and 2310 mg$Cm^{-2}d^{-1}$, with an average of 810, representing 290 mg$Cm^{-2}$, per year 45 000 tons per year of photosynthetized carbon for the whole lake. Gross photosynthetic production measured by the method of Ryther was studied over a 2-year period. The values obtained from marine water(Carry-le-Rouet) ranged from 23 to 2 337 mg$Cm^{-2}d^{-1}$, with a weighted average of 319, representing about 110 gCm$^{-2}$ per year. The values in brakish water (Etang de Berre) ranged from 14 to 1778 mg$Cm^{-2}d^{-1}$, with a weighted average of 682, representing 250 mg$Cm^{-2}$ per year and 38 400 tons per year of photosynthesized carbon for the whole lake.

• Journal of the Korean Society for Marine Environment & Energy
• /
• v.11 no.2
• /
• pp.78-85
• /
• 2008

We estimated the pollutant loads for the last 3 years based on the daily discharge at the Nakdong River dam(barrage) and spatiotemporal characteristics of seawater quality in the Nakdong river estuary to investigate the correlation between the pollutant load inflow rate and seawater quality. The main results from this research are summarized as follows. (1) The total discharge at the Nakdong River dam dam the last 11 years has been $224,576.8{\times}10^6m^3/day$. The discharge figures show that the maximum discharge occurs in August with $52,634.2{\times}10^6 m^3/day$ (23.4% of the year's volume), followed by July and Sep. in that order with 23.1 and 17%, respectively. (2) The pollutant load influx from the Nakdong River dam was composed of 307,591.3COD-kg/day, 128.746.1 TN-kg/day, and 107,625.8 TP-kg/day. (3) The surface temperature in the Nakdong River estuary was about $2.137^{\circ}C$ higher than that of the lower layer. The salinity of the lower layer was 2.209%o higher than that of the ocean surface. The salinity of the ocean surface decreased by up to 19.593%o due to the inflow of the discharge at the Nakdong River dam. (4) DO, COD, TN, and SS concentration levels tended to be higher at the ocean surface than in lower layers, whereas the reverse was true for TP. (5) The water mass at the ocean's surface and in the lower layers during the drought and flood seasons tended to be separated by the difference in densities due to the freshwater inflow.

• Korean Journal of Environmental Agriculture
• /
• v.4 no.1
• /
• pp.18-24
• /
• 1985

A pot experiment was conducted to study the influence of potassium and cesium carrier on the uptake of radionuclide $Cs^{137}$ which is an element released usually from nuclear facilities, by paddy rice upon prolonged cropping of contaminated soils. The results are summarized as follows: 1) Visual toxic symptoms on the growth of rice plant due to treatment of radioactive cesium were not observed up to $20 {\mu}Ci/10Kg$ soil in a pot. 2) The yield and potassium content in rice plant were increased with potassium application, while the reverse was true for the calcium and magnesium. The addition of potassium to the soil markedly reduced $Cs^{137}$ uptake by rice plant but the addition of Cs carrier increased $Cs^{137}$ uptake. 3) Potassium and $Cs^{137}$ showed uniform distribution in all parts of plant and the contents of these two elements were high in the stems and leaves, and low in the heads. The ratio of $Cs^{137}$ to K was, however, not uniform in all parts of a plant. It was shown that this ratio was higher in the seed part, that is, chaff and hulled grain than in the leaves and stems. 4) $Cs^{137}$ absorption rate in rice plant was remarkably reduced with increase of potassium application and it was ranged from $0.02{\sim}0.47%$ in potassium non-treated plot to 0.01∼0.04% in plot treated with a concentration of 16Kg/10a. 5) The amount of $Cs^{137}$ and potassium uptake of rice plant depended on soil type. Uptake of $Cs^{137}$ by rice plant was higher in the soil with low pH and potassium content. The $Cs^{137}$ uptake by rice plant decreased as the potassium content and pH of soil was increased, but $Cs^{137}$ uptake increased when CEC and clay content in soil was high.

• Korean Journal of Poultry Science
• /
• v.37 no.4
• /
• pp.361-372
• /
• 2010

Serological evaluation for the poultry is important for various reasons, such as designing and assessing the vaccination program and diagnosing diseases and for this reason, serologic tests for the layer flocks have been conducted on a regular basis. Moreover, the nationwide serological survey and analysis are essential to understand the epidemiological status of national poultry industry. In this sense, the study was conducted to evaluate the immune status of the layer flocks with the sera submitted to Avian Disease Laboratory, Chungbuk National University in 2009, and several important viral diseases were selected for evaluation including low pathogenic avian influenza (LPAI), Newcastle disease (ND), infectious bronchitis (IB) and avian metapneumovirus (aMPV). For LPAI and ND, the age-related patterns of geometric mean titer (GMT) changes were similar but there were differences in the flock positive rate and the level of GMT due to the different vaccination policy. In the case of IB, the values of GMT showed that the field infection was more prevalent than expected. For aMPV, positive birds in a flock increased as the layers got older, which reflected the course of field infection because vaccination against aMPV was not allowed in 2009. From this study, the immune status for the main viral diseases in layers became more clarified but this information was limited because of only one year study. Therefore, serological survey needs to be conducted on a yearly basis and furthermore include broilers and breeders for a better understanding of the health status in the national poultry industry.

• Journal of Plant Biotechnology
• /
• v.45 no.2
• /
• pp.110-116
• /
• 2018

This study was conducted to develop a simple, rapid, and reliable method for in vitro propagation of disease-free and true-to-type clones from sucker explants of thornless blackberry (Rubus fruticosus L. ${\times}$ R. parvifolius L.). To induce multiple shoots, the sucker explants were sterilized in 1% NaOCl solution, and then were aseptically cultured on the full and 1/2 MS solid medium supplemented with BAP (0.1, 0.5, 1.0, 2.0 mg/L). After six weeks of culture, the highest frequency (85.4%) of shoot formation from sucker explants was obtained on the full-strength MS medium with 1.0 mg/L BAP. Node explants obtained from multiple shoots were cultured on the various media of full- or half-strength of AD, B5, MS, SH, QL, WPM media, respectively. After 30 days of culture, plant growth was good on the half-AD, half-QL medium. After 90 days of culture, plant growth was good on the full MS and full SH medium. The survival rate of the plantlets after transfer to plastic pots containing soil mixture (sand: soil: vermiculite was 1:1:1, vol.) in the greenhouse was 98%. The results indicate that a multiple-shoot procedure can be applied for an efficient mass propagation of Rubus fruticosus L. ${\times}$ R. parvifolius L.