• Child Health Nursing Research
• /
• v.2 no.1
• /
• pp.93-111
• /
• 1996

The purpose of the study is to identify the lived experience of mothers who have children with cerebral palsy in order to understand their agony. Moreover, the result of study was to find some nursing intervention for disabled children and their mothers. For this purpose, ten mothers who are willing to cooperate with this research were selected at random from those who have children with the cerebral palsy, currently using the municipal facilities for the handicapped with cerebral malfunction. Data collection was done from October 4, 1994 th December 31, 1994. The data were collected by asking the mothers mentioned above with some unstructured open-ended questions, recorded on the tapes with permission by the interviewee in order to prevent missing of the interviewed contents. These collected data have been substantiated and properly analyzed on the basis of phenomenological approach initiated by Colaizzi's method. The results and validity are proved to be credible by means of the individual checking of the interviewed mothers. The results of this study are as follows : 1. When the mother is first informed of the diagnosis of cerebral palsy on her child, she usually misses the crucial timing needed for proper treatment of the child's disorder because she is notified through the doctor's indifference and his apparently inactive, matter-of-fact attitude. At first she suspects the doctor's diagnosis and tries to attribute it to the unknown cause from a certain genetic problem and then she quickly wants to deny the whole situation that her child is really suffering from the cerebral palsy. The reality is too much for her to accept as it is and she would not believe her child is abnormal. Therefore, she even attempts depend on the power of God for its solution. 2. The mother, who goes thorough this kind of uncommon experiences, is totally devoted to the treatment and care of the child and completely ignores her own life and happiness. At the same time, she feels sorry for her other normal children she believes having not enough care and concern. Also, she feels sorry for the sick child when the child's brothers or sisters show special concern for the patient out of sympathy. It is sorry and not satisfied for her that the child is growing with abnormality and neighbor other around have inappropriate attitudes. Likewise, she is discontent with her husband's lack of concern about the child's treatment. She believes that the health care system in this society isn't fulfilling its due purpose. In the state of her utmost distress and anxiety, she always feels the need of competent consultants, and is angry about that her child is treated as an abnormal being, she is trying to hide the child from other people and to make him or her disappear, if possible. Although she doesn't have harmonious relation with her husband, she id happy when he shows his affection for the child and she feels relieved and thankful when the relatives don't mention about the child's condition Since the child's overall status of health is continuously in unstable conditions, requiring her all-time readiness for an emergency, she feels guilty of her child's illness toward the fEmily members as if it was her own fault to have borne such an abnormal child and she feels responsible for the child morally and financially if necessary Because her life is centered on taking care of the child, she cannot afford to enjoy her own life and happiness. She is a lonely mother, fatigued, with no proper relationship with other people around her. With this sense of guilt and responsibility as a mother of an unusual disease, she has no choice but to grieve her destiny from which she is not allowed to escape. 3. Nevertheless, the mother with the child suffering from the cerebral palsy does not easily give up the hope of getting her child cured and she believes that in the long run, though slower than hoped, her abnormal son or daughter will be eventually cured to become a normal sibling someday. This kind of hope is sustained by the mother's strong faith coming from observing the progress of other similar children getting better. Sometimes she is encouraged to have this faith by other mothers who share the same painful experiences, believing that her child will improve even more rapidly than others with the same palsy. Full of hope, she painstakingly waits for the child's healing. Moreover, she plans to have another child. she thinks that the patient child's brothers and sisters only can truly understand and look after the patients. However, when she notices that the progress of other children under the treatment does not look so hopeful, she is distressed by the thoughts that her child may never get well. Too, she is worried that the patient's brother or sister will be born as the same invalid with the cerebral disease. She is discouraged to have another baby as much as she is encouraged to. She is also troubled by the thought that in case she has another baby, she will have to be forced. to neglect the patient child, especially when she does have an extra hand or some reliable person to help her with taking care of the patient.

• Journal of Korean Academy of Nursing
• /
• v.20 no.2
• /
• pp.153-173
• /
• 1990

One of the important tasks for new parents. especially mothers, is to establish warm, mutually affirming interpersonal relationships with the new baby in the family, with the purpose of promoting the healthy development of the child and the wellbeing of the whole family. Nurses assess the quality of the behavioral characteristics of the maternal-infant interaction. This study examined the relationships between primiparas pereptions of their delivery experience and their maternal infant interaction. It compared to delivery experience of mothers having a normal vaginal delivery with those having a casearean section. The purpose was to explore the relationships between the mother's perceptions of her delivery experience with her maternal infant interaction. The aim was to contribute to the development of theoretical understanding on which to base care toward promoting the quality of maternal-infant interaction. Data were collected directly by the investigator and a trained associate from Dec. 1, 1987 to March 8, 1988. Subjects were 3 random sample of 62 mothers, 32 who had a normal vaginal delivery and 30 who had a non-elective cesarean section (but without other perinatal complications) at three general hospitals in Seoul. Instruments used were the Stainton Parent -infant Interaction Scale(1981) and the Marut and Mercer Perception of Birth Scale(1979). The first observations were made in the delivery room (for vaginally delivered mothers only), followed by day 1, day 2, day 3, and 2 weeks, 4 weeks, 6 weeks and 8 weeks after birth, for a total of 7-8 contacts(Cesarean section mothers were observed on days 4 and 5 but the data not used for analysis). Observations in the hospital were made during the hour prior to scheduled feedings. The infant was placed beside the mother. Later contacts were made at home. Data analysis was done by computer using as SPSS program and indulded X² test, paired t-test, t-test, and Pearson Correlation coefficient ; the results were as follows. 1. Mothers who had a normal vaginal delivery tended to perceive the delivery experience more positively than cesarean section mothers(p＝0.002). The finding supported the hypothesis I that perception of delivery would vary according to the method of delivery. Mothers' perceptions of birth were classified into three dimensions, labor, delivery and the bady. There was a significantly different and positive perception by the vaginally delivered mothers to the delivery experience(p＝0.000) but no differences for labor or the bady according to the delivery method(p＝0.096, p＝0.389), 2. Mothers who had a normal vaginal delivery had higher average maternal-infant interaction scores(p＝0.029) than mothers who had a cesarean section. There were similar higher scores for the 1st day(p＝0.042), 2nd day (p＝0.009), and the 3rd day(p＝0.006) after delivery but not for later times. The findings supported the hypothesis Ⅱ that there would be differences in maternal-infant interaction for mothers having vaginal and cesarean section deliveries. However these differences deccreased section deliveries. However these differences decreased over time . by eight weeks the scores for vaginal delivery mothers averaged 8.1 and for cesarean section mothers, 7.9. 3. The more highly positive the pereption of the delivery experience, the higher the maternal-infant interaction score for all subjects(F＝.3206, p＝.006). The findings supported the hypothesis Ⅲ that there would be correlations between perceptions of delivery and maternal-infant interaction. The maternal infant interaction was highest when the perception of the bady and deliery was positive(r＝.4363, p＝.000, r＝.2881, p＝.012). No correlations between perceptions of labor and maternal-infant interaction were found(p＝0.062). 4. The daily maternal-infant interaction score for the initial contact after birth to 8 weeks postpartum had the lowest average score 5.20 and the highest 7.98(in a range of 0-10). This subjects group of mothers needed nursing intervention to promote their maternal- infant interaction. The daily scores for the maternal-infant over the period of eight weeks. However, there were significantly different increases in maternal-infant interaction only from the first to second day(p＝0.000) and from the fourth to sixth weeks after birth(P＝0.000). 5. When the eight items of maternal-infant interaction were evaluated separately, “Expresses feelings about her role as mother” had the highest average score, 1.64(ina range of 0-3)and “Speaks to baby” the lowest, 0.9. All items, with the possible exception of “Expresses feelings about her role as mother”, suggested the subjects' need of nursing intervention to promote maternal-infant interaction. 6. There were positive correlations between certain general charateristis, namely, both a higher economic status(p＝0.002) and breast feeding(p＝0.202) and maternal - infant interaction. There were positive correlations between a mother's confidence in her role as a mother and the perception of the birth experience(p＝0.004). For mothers who had a cesarean section, a positive perception of the birth experience was related to the duration of her marriage(p＝0.010), a wanted pregnancy (P＝0.030) and her confidence in her role as a mother(p＝0.000). Pereptions of birth for mothers who had a normal vaginal delivery were positive than those for mothers who had a cesarean section. The level of maternalinfant interaction for mothers delivered vaginally was higher than for cesarean section mothers. The relationship between perception of birth and materanalinfant interaction was confirmed. Cesarean section has an impact on the mother's perceived experience of birth which, in turn, is positively related to maternal-infant in turn, is positively related to maternal-infant interaction. Nursing intervention to enhance maternal-infant interaction should begin in prenatal classes with an exploration of the potential impact of cesarean section on the perceptions of the birth experience and continue throughout the perinatal and post-natal periods to promote the mother's ability to control with this crisis experience and to mobilize social support. Nursing should help transform a relatively negatively perceived experience into an accepted, positively perceived and self affirming experience which enhances the maternal-infant relationship.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.123-138
• /
• 2017

Since the stock market is driven by the expectation of traders, studies have been conducted to predict stock price movements through analysis of various sources of text data. In order to predict stock price movements, research has been conducted not only on the relationship between text data and fluctuations in stock prices, but also on the trading stocks based on news articles and social media responses. Studies that predict the movements of stock prices have also applied classification algorithms with constructing term-document matrix in the same way as other text mining approaches. Because the document contains a lot of words, it is better to select words that contribute more for building a term-document matrix. Based on the frequency of words, words that show too little frequency or importance are removed. It also selects words according to their contribution by measuring the degree to which a word contributes to correctly classifying a document. The basic idea of constructing a term-document matrix was to collect all the documents to be analyzed and to select and use the words that have an influence on the classification. In this study, we analyze the documents for each individual item and select the words that are irrelevant for all categories as neutral words. We extract the words around the selected neutral word and use it to generate the term-document matrix. The neutral word itself starts with the idea that the stock movement is less related to the existence of the neutral words, and that the surrounding words of the neutral word are more likely to affect the stock price movements. And apply it to the algorithm that classifies the stock price fluctuations with the generated term-document matrix. In this study, we firstly removed stop words and selected neutral words for each stock. And we used a method to exclude words that are included in news articles for other stocks among the selected words. Through the online news portal, we collected four months of news articles on the top 10 market cap stocks. We split the news articles into 3 month news data as training data and apply the remaining one month news articles to the model to predict the stock price movements of the next day. We used SVM, Boosting and Random Forest for building models and predicting the movements of stock prices. The stock market opened for four months (2016/02/01 ~ 2016/05/31) for a total of 80 days, using the initial 60 days as a training set and the remaining 20 days as a test set. The proposed word - based algorithm in this study showed better classification performance than the word selection method based on sparsity. This study predicted stock price volatility by collecting and analyzing news articles of the top 10 stocks in market cap. We used the term - document matrix based classification model to estimate the stock price fluctuations and compared the performance of the existing sparse - based word extraction method and the suggested method of removing words from the term - document matrix. The suggested method differs from the word extraction method in that it uses not only the news articles for the corresponding stock but also other news items to determine the words to extract. In other words, it removed not only the words that appeared in all the increase and decrease but also the words that appeared common in the news for other stocks. When the prediction accuracy was compared, the suggested method showed higher accuracy. The limitation of this study is that the stock price prediction was set up to classify the rise and fall, and the experiment was conducted only for the top ten stocks. The 10 stocks used in the experiment do not represent the entire stock market. In addition, it is difficult to show the investment performance because stock price fluctuation and profit rate may be different. Therefore, it is necessary to study the research using more stocks and the yield prediction through trading simulation.

• Journal of Intelligence and Information Systems
• /
• v.23 no.1
• /
• pp.47-67
• /
• 2017

Steel plate faults is one of important factors to affect the quality and price of the steel plates. So far many steelmakers generally have used visual inspection method that could be based on an inspector's intuition or experience. Specifically, the inspector checks the steel plate faults by looking the surface of the steel plates. However, the accuracy of this method is critically low that it can cause errors above 30% in judgment. Therefore, accurate steel plate faults diagnosis system has been continuously required in the industry. In order to meet the needs, this study proposed a new steel plate faults diagnosis system using Simultaneous MTS (S-MTS), which is an advanced Mahalanobis Taguchi System (MTS) algorithm, to classify various surface defects of the steel plates. MTS has generally been used to solve binary classification problems in various fields, but MTS was not used for multiclass classification due to its low accuracy. The reason is that only one mahalanobis space is established in the MTS. In contrast, S-MTS is suitable for multi-class classification. That is, S-MTS establishes individual mahalanobis space for each class. 'Simultaneous' implies comparing mahalanobis distances at the same time. The proposed steel plate faults diagnosis system was developed in four main stages. In the first stage, after various reference groups and related variables are defined, data of the steel plate faults is collected and used to establish the individual mahalanobis space per the reference groups and construct the full measurement scale. In the second stage, the mahalanobis distances of test groups is calculated based on the established mahalanobis spaces of the reference groups. Then, appropriateness of the spaces is verified by examining the separability of the mahalanobis diatances. In the third stage, orthogonal arrays and Signal-to-Noise (SN) ratio of dynamic type are applied for variable optimization. Also, Overall SN ratio gain is derived from the SN ratio and SN ratio gain. If the derived overall SN ratio gain is negative, it means that the variable should be removed. However, the variable with the positive gain may be considered as worth keeping. Finally, in the fourth stage, the measurement scale that is composed of selected useful variables is reconstructed. Next, an experimental test should be implemented to verify the ability of multi-class classification and thus the accuracy of the classification is acquired. If the accuracy is acceptable, this diagnosis system can be used for future applications. Also, this study compared the accuracy of the proposed steel plate faults diagnosis system with that of other popular classification algorithms including Decision Tree, Multi Perception Neural Network (MLPNN), Logistic Regression (LR), Support Vector Machine (SVM), Tree Bagger Random Forest, Grid Search (GS), Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The steel plates faults dataset used in the study is taken from the University of California at Irvine (UCI) machine learning repository. As a result, the proposed steel plate faults diagnosis system based on S-MTS shows 90.79% of classification accuracy. The accuracy of the proposed diagnosis system is 6-27% higher than MLPNN, LR, GS, GA and PSO. Based on the fact that the accuracy of commercial systems is only about 75-80%, it means that the proposed system has enough classification performance to be applied in the industry. In addition, the proposed system can reduce the number of measurement sensors that are installed in the fields because of variable optimization process. These results show that the proposed system not only can have a good ability on the steel plate faults diagnosis but also reduce operation and maintenance cost. For our future work, it will be applied in the fields to validate actual effectiveness of the proposed system and plan to improve the accuracy based on the results.

• The Korean Journal of Mycology
• /
• v.21 no.3
• /
• pp.200-211
• /
• 1993

Somatic hybrids of Pleurotus florida ASI 2016 and Pleurotus ostreatus ASI 2018 were obtained by protoplast fusion. The 40 fusants($P1{\sim}P40$) was examined for the yield on fermented and pasteurized rice straw in a tray. The carpophore yield of them were showed as the range of $27.0{\sim}155.2$, based on parental values of 100(ASI 2018), The pilei of fusants between orange white colored P. florida and dark grey colored P. ostreatus had mixed colors in the young stage. Other breeding programmes were performed to improve new varieties with high yield and good quality. A new oyster mushroom variety, Wonhyeongneutaribeosus(P72), was developed at the Agricultural Sciences Institute, Rural Development Administration in 1990. This P.florida-ostreatus-ostreatus hybrid P72 was selected from 38 protoplast fusion products($P41{\sim}P78$) between P.florida-ostreatus recombinant P5-M 43-arg rib and P. ostreatus ASI 2-13-0 2001-19-pro orn. The yield indexes of 38 hybrids ranged $40.5{\sim}152.7$ compared with the parental values of 100(ASI 2001). Hybrid P72 was characterized by the large fruiting bundle of semispherical shape with long stipe and by the small and circular pileus, resulting in lower harvesting cost. A significant increase in carpophore production was observed in somatic hybrids of protoplasts due to heterosis. A comparision of hybrid with parents P72 was made using isozyme analysis. The esterase banding patterns could be characterized by new bands in the hybrids. Seven fusion products of four crosses between P.florida ASI 2016 and P. ostreatus ASI 2018 were analysed with respect to the distribution of progenies and segregation of gene markers by random basidiospore analysis. Segregation of alleles should yield progeny of four genotypes in a Mendelian ratio of 1 : 1 : 1 : 1 for prototrophs, auxotrophs of one parental type, auxotrophs of the other parental type, and auxotrophic recombinants, respectively. However, five fusants of them did not detect one parental, P.ostreatus, type. Basidiospores could yield progeny of 16 genotypes in the cross of one of the recombinant P5-M43-arg $rib{\times}P. ostreatus$ ASI 2-13-pro orn but the segregants of three fusants were not detected clearly. The allele ratio of loci could be expected 1 : 1 : 1 : 1 for arg, rib, pro and orn. The ratio, however, would be changed to 4 : 1 : 1 : 1 with increasing proportion of argo In almost all the fusants, prototrophic recombinants were recovered in large numbers against auxotrophic markers. Parental genotypes were recovered with the recombinant progeny amounting to $38.68{\sim}99.56%$. The analysis provides proof of heterokaryosis and strong evidence for haploidy of vegetative nuclei, a sexual cycle consisting of nuclear fusion and meiosis.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• v.6 no.1
• /
• pp.56-64
• /
• 1995

Family is a primary unit of the major socialization processing for children. Parents among the family members are one of the most important figures from whom the child and adolescent acquire a wide variety of behavior patterns, attitudes, values and norms. An organization of family members product family structural functioning. Abnormal family structure is one of the most important reference models in the learning of antisocial patterns of behavior. Therefore incest and child sexual abuse including spouse abuse, elderly abuse, and neglect occurs in the abnormal family structural setting. In particular, incest, a specific form of sexual abuse, was once thought to be a phenomenon of great rarity, but our clinical experiences, especially over the past decade, have made us aware that incest and child sexual abuse is not rare case and on the increasing trend. Therefore, the aim of this study was to determine the family problem and dynamics of incest family, and character pattern of post-incest adolescent victim in Korea. A total of 1,838 adolescents from middle and high school(1,237) and juvenile correctional institute(601) were studied, sampled from Korean student population and adolescent delinquent population confined in juvenile correctional institutes, using proportional stratified random sampling method. The subjects' ages ranged from 12 to 21 years. Data were collected through questionnaire survey. Data analysis was done by IBM PC of Behavior Science Center at the Korea university, using SAS program. Statistical methods employed were Chi-square, principal component analysis and t-test etc. The results of this study were as follows ; 1) Of 1,071 subjects, 40(3.7%) reported incest experiences(sibling incest : 1.6% ; another type of incest : 2.1%) in their family setting. 2) The character pattern of post-incest adolescent victim was more socially maladjusted, immature, impulsive, rigid, anxious and dependent than non-incest adolescent. Also they showed some problem in academic performance and their assertiveness. 3) The other family members of incest family revealed more psychological and behavioral problem such as depression, alcoholism, psychotic disorder and criminal act than the non-incest family, even though there is no evidence of the context between them. 4) The family dynamics of incest family tended to be dysfunctional trend, as compared with non-incest family. It showed that the psychological instability of family member, parental rejection toward their children, coldness and indifference among family member and marital discordance between the parents had significant correlation with incest.

• Journal of Nutrition and Health
• /
• v.7 no.3
• /
• pp.1-13
• /
• 1974

The relative state of human iron storage may be ascertained more reliably through determination of the serum iron, iron binding capacity, transferrin saturation and absorption of radioactive iron in conjunction with studies of red cell morphology than from the study of red cell morphology alone. Recent investigations have shown that there is an increase in red cell protoporphyrin concentration in iron deficiency anemia. The significance of the red cell protoporphyrin has been discussed greatly during the years since its discovery. Two of the main factors which appear to influence the amaunt of protoporphyrin are increased erythropoiesis and factors interfering with the utilization of iron in the synthesis of hemoglobin, and iron deficiency. Recently Heller et al. have described a simplified method for blood protoporphyrin assay and this technique could be used assess nutritional iron status, wherein even minor insufficiencies are detectable as increased protoporphyrin concentrations. Based on the evaluation of the relationship between nutritional iron status and red cell protoporphyrin as an index suitable for the detection of the iron deficiency is described in this paper. RESULTS 1. Hemoglobin Concentrations and Anthropometric Measurements. The mean and standard deviations of the various anthropometric measurements of different age and sex groups are shown in table 1. There measurements have been compared with the Korean Standard. In the absence of local standards for arm circumference and skin-fold thickness over triceps, they have been compared with the standard from Jelliffe. Table 2,3, and 4 give anthropometric measurements and frequency (%) of anemia in children surveyed. The mean height of the children studid was 10 to 20 percent; below the Korean Standard. The distribution of height below 80 percent of the Standard was 21.2 percent, however, among anemic group this percentage was 27.7 percent. In general, the mean weight of the children was 10 to 15 percent below the Korean Standard. The percentage of children with weight less than 80 percent of the Standard was about 35 percent. But in the anemic group of the children, this percentage was 44 percent. The mean arm circumference was about 15 percent lower than the Jelliffe's standard. 61.2 percent of the children had values of arm circumference below 80 percent of the standard. Children with low hemoglobin levels, this percentage was 80 percent. The mean skinfold thickness over the triceps of the children studied was about 25 Percent lower than the Jelliffe's standard and 61.2 percent of the children had the value less than 80 percent of the standard. Among anemic children, this percentage was 70.8%. As may be seen from table 5, the mean hemoglobin concentration of the total group was 11.3g/100ml. Hemoglobin concentration was less than 11.0g/100ml. in 65(36.5%) of the 178 children. The degree of anemia in most of these children was mild with a hemoglobin level of less than 8.0g/100ml. found in only one child. In general, the prevalence of anemia was high in female children than male and decreased its frequency with increasing age. Relatively close relationship was observed between hemoglobin level and anthrophometric measurements especially high between arm circumference and skinfold thickness and hemoglobin but very low in height and low in weight and hemoglobin level, estimated by chi-square value. II. Serum iron, Transferrin saturation (1) Serum iron, and transferrin saturation Serum iron, transferrin saturation and red cell protoporphyrin concentrations were estimated in sub-sample of 84 children from 1 to 6 years and 24 older children between 7 and 13 years of age. The findings are presented in table 6. The mean serum iron concentration of the total group was 59ug/100ml. However, the level incrased with age from 36.6ug/100ml. (1-3years) to 80.8ug/100ml. (7-13 years). 60 percent of these children had a serum iron level less than 50ug/10ml. in the 1-3 years age group and 31.4 percent for 4-6 years group. These contrast with the finding of 12.5 percent anemic children in the 7-13 years age group. The mean transferrin saturation for the total group was 18.1 percent and frequency of anemia by transferrin saturation was observed same pattern as serum iron concentration. (2) Red cell protoporphyrin concentrations. (a) Red cell protoporphrin levels of children: Red cell protoporphyrin and other biochemical data are shown in table 4. The mean concentration in red cell of all children was fround 46.3ug/100ml. RBC. and differences with age groups were observed; in the age group 1-3 years, the mean concentration was $59.5{\pm}32.14$ ug/100ml. RBC; 4-6 years $44.1{\pm}22.57$ ug/100ml. RBC. and 7-13 years, $39.0{\pm}13.56$ ug/100ml. RBC. (b) Normal protoporphyrin values in adults: It was observed that in 10 normal adult males studied here the level of protoporphyrin in red cell ranged from 18 to 54 ug/100ml. RBC. and the mean concentration was $47.5{\sim}14.47$ ug/100ml. RBC. Other biochemical determination made on the same subjects are presented in table 8. (c) Red tell protoporphyrin concentration of occupational blood donors: The results of analyses for red cell protoporphyrin as well as serum iron, transferrin saturation and hemoglobin in the 76 blood donors are presented in table 7 and 8. In this experiment, donors were selected at random, however, most of them bled repeatedly because of poor economic situation, I doubt. Table 9 shows the distribution of red cell protoporphyrin concentration and hemoglobin concentration of occupational donors. The mean hemoglobin value for the total was 11.9 g/100 ml. When iron deficiency anemia is defined as a transferrin saturation below 15%, prevalence of anemia was 47.4 percent and the mean serum iron was 27.1ug/100ml. and red cell protoporphyrin, 168.3ug/100ml. RBC. However, mean serum iron and protoporphyrin concentration of above 15% transferrin saturation were 11.6 ug/100 ml. and 58.8 ug/100 ml. RBC. respectively. The mean Protoporphyrin concentration of non-anemic (above 15% transferrin saturation) donors was slightly higher than the results of normal adult males.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.167-181
• /
• 2018

Over the past decade, deep learning has been in spotlight among various machine learning algorithms. In particular, CNN(Convolutional Neural Network), which is known as the effective solution for recognizing and classifying images or voices, has been popularly applied to classification and prediction problems. In this study, we investigate the way to apply CNN in business problem solving. Specifically, this study propose to apply CNN to stock market prediction, one of the most challenging tasks in the machine learning research. As mentioned, CNN has strength in interpreting images. Thus, the model proposed in this study adopts CNN as the binary classifier that predicts stock market direction (upward or downward) by using time series graphs as its inputs. That is, our proposal is to build a machine learning algorithm that mimics an experts called 'technical analysts' who examine the graph of past price movement, and predict future financial price movements. Our proposed model named 'CNN-FG(Convolutional Neural Network using Fluctuation Graph)' consists of five steps. In the first step, it divides the dataset into the intervals of 5 days. And then, it creates time series graphs for the divided dataset in step 2. The size of the image in which the graph is drawn is $40(pixels){\times}40(pixels)$, and the graph of each independent variable was drawn using different colors. In step 3, the model converts the images into the matrices. Each image is converted into the combination of three matrices in order to express the value of the color using R(red), G(green), and B(blue) scale. In the next step, it splits the dataset of the graph images into training and validation datasets. We used 80% of the total dataset as the training dataset, and the remaining 20% as the validation dataset. And then, CNN classifiers are trained using the images of training dataset in the final step. Regarding the parameters of CNN-FG, we adopted two convolution filters ($5{\times}5{\times}6$ and $5{\times}5{\times}9$) in the convolution layer. In the pooling layer, $2{\times}2$ max pooling filter was used. The numbers of the nodes in two hidden layers were set to, respectively, 900 and 32, and the number of the nodes in the output layer was set to 2(one is for the prediction of upward trend, and the other one is for downward trend). Activation functions for the convolution layer and the hidden layer were set to ReLU(Rectified Linear Unit), and one for the output layer set to Softmax function. To validate our model - CNN-FG, we applied it to the prediction of KOSPI200 for 2,026 days in eight years (from 2009 to 2016). To match the proportions of the two groups in the independent variable (i.e. tomorrow's stock market movement), we selected 1,950 samples by applying random sampling. Finally, we built the training dataset using 80% of the total dataset (1,560 samples), and the validation dataset using 20% (390 samples). The dependent variables of the experimental dataset included twelve technical indicators popularly been used in the previous studies. They include Stochastic %K, Stochastic %D, Momentum, ROC(rate of change), LW %R(Larry William's %R), A/D oscillator(accumulation/distribution oscillator), OSCP(price oscillator), CCI(commodity channel index), and so on. To confirm the superiority of CNN-FG, we compared its prediction accuracy with the ones of other classification models. Experimental results showed that CNN-FG outperforms LOGIT(logistic regression), ANN(artificial neural network), and SVM(support vector machine) with the statistical significance. These empirical results imply that converting time series business data into graphs and building CNN-based classification models using these graphs can be effective from the perspective of prediction accuracy. Thus, this paper sheds a light on how to apply deep learning techniques to the domain of business problem solving.

• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.1-19
• /
• 2018

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image can be classified during fashion editorial task to classify and label the brand or style, and website query image can be processed by e-commerce multi-complex service providing item information or recommending similar item.

• Journal of Global Scholars of Marketing Science
• /
• v.17 no.4
• /
• pp.23-49
• /
• 2007

Salespeople act at the point of MOT with customers and deliver the enterprise's message to the customers. They build up relationships with customers as well as deliver the customer's message to the enterprise. The salespeople's activity at the point of MOT with the customers and the degree of satisfaction of the customers' needs will affect the customers' attitude toward the enterprise, brand loyalty, and retention intention. Ultimately, it will influence the enterprise's financial performance. The control of salespe1ople is one of the most interesting topics of marketing. This research investigates the relationships of the control types over salespeople(positive/negative outcome control, positive/negative behavior control) and job satisfaction and their mediating variables. The mediating variables in the relationships have been identified as outcome/behavior-related role clarity and self-efficacy. The purpose of this study is more specifically as follows: First, it investigate how the perception of salespeople control types affect role-clarity. Second, it examines how the perception of salespeople control types influence self-efficacy. Third, it investigate the mediating role of role-clarity between the perception of salespeople control types and self-efficacy. Fourth, it investigates how role-clarity affect self-efficacy and job satisfaction. Finally, it will investigates how self-efficacy influences job satisfaction. Data were collected from the pharmaceutical industry salespeople and analyzed by SPSS 12.0 and AMOS　6.0. The data were collected by 400 respondents and 377 valid questionnaires were analyzed. The results are summarized as follows: First, positive/negative outcome controls had a positive relationship with outcome-related role clarity. Also positive behavior control had a positive effect on behavior-related role clarity, but negative behavior control didn't influence behavior-related role clarity. Second, positive outcome control influenced self-efficacy positively, but positive behavior control didn't have a positive effect on self-efficacy. In addition negative outcome control and negative behavior control had a positive effect on self-efficacy due to the mediating role of outcome-related and behavior-related role clarity. Third, outcome-related role clarity and behavior-related role clarity influenced self-efficacy positively. Behavior-related role clarity had a positive effect on job satisfaction, but outcome-related role clarity didn't influence job satisfaction. Finally, self-efficacy didn't have any effect on job satisfaction. The contributions of this study are as follows: First, existing studies have investigated the direct causal relationship between salespeoples' control type and performance, but this study investigates the structural causality between salespeoples' control types, responses, and performances. Second, this study found the mediating role of outcome-related/behavior-related role-clarity between outcome/behavior control and self-efficacy. Finally, the findings of this study further insight to existing studies on the relationship between job satisfaction and self-efficacy. The confidence of salespeoples' task influenced job satisfaction positively in existing articles,field studies, but the relationship between these two variables was not significant in this study. This means that there can be a different relationship between confidence and job satisfaction according to salespeoples' business. That is, the business environment may not be satisfying, even if the salespeople say that they have ability and confidence about their business. This means that able salespeople who have ability and confidence about their business are not satisfied with their job advancement in the company. Therefore, enterprise need to provide training that can establish a business environment that can satisfy the salespeole's expectation level which will secure good salespeople. This study may have limitation when applied to future studies. First,in this study as with existing studies it investigates the control level that salespeople feel is being measured. Actuality, the control level that a manager enforces and the control level that salespeople perceive when one is late can be different. There is need to measure lateness from both the perspective of the manager and salespeople should be done to supplement this study in the future Second, this study used variables that were connected with action result but salespeople's job satisfaction is due to the result of control. But, focusing on result of control can provide a more important financial result than sales performance. This study is also limited in that it did not consider financial result by result of control. Further studies on this will need to be done in the future. Third, this study may have a further limitation,because the investigation was restricted to pharmaceutical salespeople selling to hospitals. It is necessary to execute investigations in various industries to increase the generalization of the study findings Fourth, in this study, role clarity and self-efficacy by response variable for control and considered job satisfaction by outcome variable of control was considered. But, can other variables be considered beside response variable and result variable for control? For example, can financial affairs and change of post by outcome variable along with business stress by response variable for control be considered? Therefore, future studies need to consider various control variables. Finally, there is limited supporting research in the field of marketing which restricts the generalization of the study finding along with collecting material through random sampling of a limited size. This research summarizes the research in this area, the difference from the previous research, and provides a discussion of its limitations and the need and direction for further future research.