• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.47-73
• /
• 2020

KTX rolling stocks are a system consisting of several machines, electrical devices, and components. The maintenance of the rolling stocks requires considerable expertise and experience of maintenance workers. In the event of a rolling stock failure, the knowledge and experience of the maintainer will result in a difference in the quality of the time and work to solve the problem. So, the resulting availability of the vehicle will vary. Although problem solving is generally based on fault manuals, experienced and skilled professionals can quickly diagnose and take actions by applying personal know-how. Since this knowledge exists in a tacit form, it is difficult to pass it on completely to a successor, and there have been studies that have developed a case-based rolling stock expert system to turn it into a data-driven one. Nonetheless, research on the most commonly used KTX rolling stock on the main-line or the development of a system that extracts text meanings and searches for similar cases is still lacking. Therefore, this study proposes an intelligence supporting system that provides an action guide for emerging failures by using the know-how of these rolling stocks maintenance experts as an example of problem solving. For this purpose, the case base was constructed by collecting the rolling stocks failure data generated from 2015 to 2017, and the integrated dictionary was constructed separately through the case base to include the essential terminology and failure codes in consideration of the specialty of the railway rolling stock sector. Based on a deployed case base, a new failure was retrieved from past cases and the top three most similar failure cases were extracted to propose the actual actions of these cases as a diagnostic guide. In this study, various dimensionality reduction measures were applied to calculate similarity by taking into account the meaningful relationship of failure details in order to compensate for the limitations of the method of searching cases by keyword matching in rolling stock failure expert system studies using case-based reasoning in the precedent case-based expert system studies, and their usefulness was verified through experiments. Among the various dimensionality reduction techniques, similar cases were retrieved by applying three algorithms: Non-negative Matrix Factorization(NMF), Latent Semantic Analysis(LSA), and Doc2Vec to extract the characteristics of the failure and measure the cosine distance between the vectors. The precision, recall, and F-measure methods were used to assess the performance of the proposed actions. To compare the performance of dimensionality reduction techniques, the analysis of variance confirmed that the performance differences of the five algorithms were statistically significant, with a comparison between the algorithm that randomly extracts failure cases with identical failure codes and the algorithm that applies cosine similarity directly based on words. In addition, optimal techniques were derived for practical application by verifying differences in performance depending on the number of dimensions for dimensionality reduction. The analysis showed that the performance of the cosine similarity was higher than that of the dimension using Non-negative Matrix Factorization(NMF) and Latent Semantic Analysis(LSA) and the performance of algorithm using Doc2Vec was the highest. Furthermore, in terms of dimensionality reduction techniques, the larger the number of dimensions at the appropriate level, the better the performance was found. Through this study, we confirmed the usefulness of effective methods of extracting characteristics of data and converting unstructured data when applying case-based reasoning based on which most of the attributes are texted in the special field of KTX rolling stock. Text mining is a trend where studies are being conducted for use in many areas, but studies using such text data are still lacking in an environment where there are a number of specialized terms and limited access to data, such as the one we want to use in this study. In this regard, it is significant that the study first presented an intelligent diagnostic system that suggested action by searching for a case by applying text mining techniques to extract the characteristics of the failure to complement keyword-based case searches. It is expected that this will provide implications as basic study for developing diagnostic systems that can be used immediately on the site.

• Journal of the Korean Society of Costume
• /
• v.30
• /
• pp.261-274
• /
• 1996

In the present age dominate by both cer-taingty of 1% and uncertainty of 99% 'Fuzzy thinking' of Bart Kosko that is the way to solve the problems by the scientific way through a worldview of Buddhism or Taoism greatly prevails around the world over 'Lateral greatly prevails around the world over 'Lateral thinking' and the authenticity or the right and-wrong of the uncertainty which is the thinking way to find the answer of the problems of illogical way of Edward de Beno against the western vertical thinking were Concurrently fashion designers over the world also adopt the oriental elements. But there exist differences of thoughts between the orient and the occident. And they have dif-ferent thinking way of aestheticism and references of the value on the beauty. Not only beauty but the view through the mind as intuitional thought in which not only the rec-ognition of sense but also the rationalism and the naturalness play key role. The aesthetic sense in the orient contains both the truth and virtue. 2) The beauty of the mean It's from the thought of neutralization of Confucius. The mean or moderation state which in harmony with ethical virtue and aes-thetic beauty is the ideal and is the ultimate. Therefore the thought of Confucian is the creativity in which the balance and the har-mony is most important. Fashion design is also one of the representation of the mean because the spirit of the designer is harmonized for-mlessly with the object of the model of the fashion design. 2) The beauty of skillfulness It indicates the Taoism of Lao-tzu and Chuangtzu. It takes a super-artistic declar-ation that human can feel and recognize the color of colorlessness the sound of sound-lessness and the taste of tastelessness. The thought of arts affected by Taoism is 'ad-vanced age' called the beauty of skillfulness. The view of arts of lao-tzu takes the beauty of cosmos and the nature as a standard. Es-pecially the beauty of inactivity is recognized by the linkage between the beauty and the ugliness. And these things appear in fashion design as a design element such as humor or exaggeration. 3) The beauty of non-dualism It is thought of Buddhism that all evil passions of worry occur form the opposition in dualism. Finally this thought leads to that everything is consistent and truth is only one from the point of view that virtue and vice has on linkage that is 'no virtues no vices' and 'one with two, two, with one, one is not two' A big tendency like this became the root forma-tion of the thought of the oriental arts. 3. Characteristics of the oriental aesthetic sense on the present fashion design 1) The formation of the fashion design on the oriental elements In the picture-incantation which was a representation of an era when the thought of 'cosmic dual forces' dominated the basic polygons of 'a circle square triangle' means both 'one two three' and 'the negative positive mean' of cosmic elements. From this point of view the was of planner cutting in the Orient is dif-ferent from that of the Occidental which is in three-dimensional. The planner polygon type of the cut-pieces comes to have the meaning of the three-dimension when they consist of a suit that has the combination of each cut-piece. This shows the consistency with the principle of cosmos creation of Taoism that one is two two is three and three is every-thing. 2) The coloring and the symbolic represen-tation of the fashion design on the orien-tal elements The sense on the colors in the Orient from the thought of 'the cosmic dual forces and the five elements' is not the experi-enced from the knowledge but contains the consideration of philosophy Five-primary-color representing compass directions Blue(East) Red(South) Yellow(Center) White (West) and Black (North) is called ' the posi-tive' for this five-primary-color secondary-color which comes from the compound of the primary colors is called 'the negative' The thought of 'the cosmic dual forces and the five elements' is also an theory containing the natural order of the cos-mos and this shows the perceptional differ-ence that they are not conceptual but to be recognized and fell directly. A thought of Buddhism which is 'Colors are colorlessness and Clolorlessness are color's proves that. 3) The pattern and symbolic representation of the fashion design on the oriental elements The pattern as a visual style is a figure of symbolic representation which adopt the mental and physical world of human and are the compo-sition of artistic revelation of the human nature and the religous thought of incantation. Es-pecially the symbolic representation of the oriental thought of Confusion. Buddhism and Taoism There are patterns such as plants aminals the oriental four gods and geometry. From the above it's the time toward the 21'th century when the world is constructing one global area and one historical zone. And the exotic mood of the Orient represented in the fashion which doesn't make the common feeling in general does not cease to develop only to express the visual modeling but also adopts the thought religion and the art which are the root of the Orientail and contains inherent willing of modeling.

• Journal of Intelligence and Information Systems
• /
• v.17 no.4
• /
• pp.227-240
• /
• 2011

New concepts and ideas often result from extensive recombination of existing concepts or ideas. Both researchers and developers build on existing concepts and ideas in published papers or registered patents to develop new theories and technologies that in turn serve as a basis for further development. As the importance of patent increases, so does that of patent analysis. Patent analysis is largely divided into network-based and keyword-based analyses. The former lacks its ability to analyze information technology in details while the letter is unable to identify the relationship between such technologies. In order to overcome the limitations of network-based and keyword-based analyses, this study, which blends those two methods, suggests the keyword network based analysis methodology. In this study, we collected significant technology information in each patent that is related to Light Emitting Diode (LED) through text mining, built a keyword network, and then executed a community network analysis on the collected data. The results of analysis are as the following. First, the patent keyword network indicated very low density and exceptionally high clustering coefficient. Technically, density is obtained by dividing the number of ties in a network by the number of all possible ties. The value ranges between 0 and 1, with higher values indicating denser networks and lower values indicating sparser networks. In real-world networks, the density varies depending on the size of a network; increasing the size of a network generally leads to a decrease in the density. The clustering coefficient is a network-level measure that illustrates the tendency of nodes to cluster in densely interconnected modules. This measure is to show the small-world property in which a network can be highly clustered even though it has a small average distance between nodes in spite of the large number of nodes. Therefore, high density in patent keyword network means that nodes in the patent keyword network are connected sporadically, and high clustering coefficient shows that nodes in the network are closely connected one another. Second, the cumulative degree distribution of the patent keyword network, as any other knowledge network like citation network or collaboration network, followed a clear power-law distribution. A well-known mechanism of this pattern is the preferential attachment mechanism, whereby a node with more links is likely to attain further new links in the evolution of the corresponding network. Unlike general normal distributions, the power-law distribution does not have a representative scale. This means that one cannot pick a representative or an average because there is always a considerable probability of finding much larger values. Networks with power-law distributions are therefore often referred to as scale-free networks. The presence of heavy-tailed scale-free distribution represents the fundamental signature of an emergent collective behavior of the actors who contribute to forming the network. In our context, the more frequently a patent keyword is used, the more often it is selected by researchers and is associated with other keywords or concepts to constitute and convey new patents or technologies. The evidence of power-law distribution implies that the preferential attachment mechanism suggests the origin of heavy-tailed distributions in a wide range of growing patent keyword network. Third, we found that among keywords that flew into a particular field, the vast majority of keywords with new links join existing keywords in the associated community in forming the concept of a new patent. This finding resulted in the same outcomes for both the short-term period (4-year) and long-term period (10-year) analyses. Furthermore, using the keyword combination information that was derived from the methodology suggested by our study enables one to forecast which concepts combine to form a new patent dimension and refer to those concepts when developing a new patent.

• Child Health Nursing Research
• /
• v.6 no.2
• /
• pp.224-239
• /
• 2000

This descriptive study was conducted to understand the contents and degree of parental stress level in the NICU patients, and to give a baseline data in developing nursing intervention program. Subjects were the 62 mother of hospitalized newborn in NICU of 1 University Hospital in Taejon City from May 1st, 1999 to November 30th, 1999, who agreed to take part in this study. The instrument used in this study were Parental Stressor Scales : NICU(PSS:NICU) developed by Miles et al. and validated by 3 NICU practitioners and 3 child health nursing faculties. The questionnaire has 4 dimensions and 45 items; sight and sounds of NICU(5 items), babies' appearance and behavior(19 items), parental role alteration and relationship with their baby(10 items), communication with health team(11 items). The questionnaire asks parents to rate each item on a five-point Likert type scale that ranges from (1) to (5). Total scores representing overall stress from the NICU environment are calculated by summing response to each item. A high score indicates high stress. A subscale score is calculated by summing the responses to each item in the subscale. Cronbach's α coefficients were .93. The data was analyzed as average, Frequency, Standard deviation, t-test, ANOVA, Pearson correlation coefficient by use of SPSS/PC+. The results of this study is summarized as follows ; 1. The total perceived stress level score of mothers was slightly high(3.6±.7). The highest scored dimension was 'appearance and behavior of the baby'(3.9±1.5), and next were 'relationship with their baby and parental role change'(3.5±1.4), 'communication with health team'(3.4±.9), 'sight and sounds of NICU'(3.2±.8). 2. Two variables were statistically significant with PSS:NICU total scale ; mother's perceived severity of the baby's condition (r=.482, P=.002) and mother's religious attendance(t=2.83, P=.01). The more the mother perceive their baby's condition severe, the higher the total stress score. There were high stress score noted in the mother of no religious attendance. 3. Four variables were statistically significant with NICU environment subscale ; mother's educational background(F=3.45, P=.04), religious attendance(t=2.28, P=.04), sex of the baby(t=2.83, P=.01) and NICU patients' hospital day(r=.359, P=.004). That is mother with high educational background and girl baby were high NICU environment subscale score. 4. Four variables were statistically significant with appearance and behavior of the baby subscale ; when first saw baby(F=3.52, P=.04), incubator care(t=2.83, P=.01), mother's perceived severity of the baby's condition(r=.303, P=.017), number of NICU visit(r=.441, P=.002). That is, seeing the baby first in the NICU and recieved incubator care was very stressful. Also, the more the mother perceive their baby's condition severe and more NICU visit, the higher the appearance and behavior of the baby subscale stress score. 5. Four variables were statistically significant with relationship with their baby and parental role change subscale ; when first saw baby(F=3.37, P=.04), sex of the baby(t=2.36, P=.03), incubator care(t=5.60, P=.00), mother's perceived severity of the baby's condition(r=.401, P=.001). That is, seeing the baby first in the NICU and girl baby was very stressful. Also, the more the mother perceive their baby's condition severe, the higher the relationship with their baby and parental role change subscale stress score. 6. Three variables were statistically significant with communication with health team subscale ; mother's educational background (F=3.63, P=.04), incubator care(t=4.24, P=.00), gestational age(r=-.394, P=.047), and birth weight(r=-.460, P=.004). That is, mother with high educational background and receiving incubator care were high communication with health team subscale score. Also, the shorter the gestational age and smaller the baby's birth weight, the higher the communication with health team subscale score. In conclusion, information about physical environment of NICU, the mother's perceived severity of baby's illness state, maternal role change related variables and the knowledge of characteristics of NICU patients must be included in nursing intervention program of mother's of NICU patients in reducing the maternal stress and anxiety level.

• Asia pacific journal of information systems
• /
• v.19 no.3
• /
• pp.99-124
• /
• 2009

Many firms in Korea have adopted and used advanced information technology in an effort to boost efficiency. The process of adapting to the new technology, at the same time, can vary from one firm to another. As such, this research focuses on several relevant factors, especially the roles of social interaction as a key variable that influences the technology adaptation process and the outcomes. Thus far, how a firm goes through the adaptation process to the new technology has not been yet fully explored. Previous studies on changes undergone by a firm or an organization due to information technology have been pursued from various theoretical points of views, evolved from technological and institutional views to an integrated social technology views. The technology adaptation process has been understood to be something that evolves over time and has been regarded as cycles between misalignments and alignments, gradually approaching the stable aligned state. The adaptation process of the new technology was defined as "appropriation" process according to Poole and DeSanctis (1994). They suggested that this process is not automatically determined by the technology design itself. Rather, people actively select how technology structures should be used; accordingly, adoption practices vary. But concepts of the appropriation process in these studies are not accurate while suggested propositions are not clear enough to apply in practice. Furthermore, these studies do not substantially suggest which factors are changed during the appropriation process and what should be done to bring about effective outcomes. Therefore, research objectives of this study lie in finding causes for the difference in ways in which advanced information technology has been used and adopted among organizations. The study also aims to explore how a firm's interaction with social as well as technological factors affects differently in resulting organizational changes. Detail objectives of this study are as follows. First, this paper primarily focuses on the appropriation process of advanced information technology in the long run, and we look into reasons for the diverse types of the usage. Second, this study is to categorize each phases in the appropriation process and make clear what changes occur and how they are evolved during each phase. Third, this study is to suggest the guidelines to determine which strategies are needed in an individual, group and organizational level. For this, a substantially grounded theory that can be applied to organizational practice has been developed from a longitudinal comparative case study. For these objectives, the technology appropriation process was explored based on Structuration Theory by Giddens (1984), Orlikoski and Robey (1991) and Adaptive Structuration Theory by Poole and DeSanctis (1994), which are examples of social technology views on organizational change by technology. Data have been obtained from interviews, observations of medical treatment task, and questionnaires administered to group members who use the technology. Data coding was executed in three steps following the grounded theory approach. First of all, concepts and categories were developed from interviews and observation data in open coding. Next, in axial coding, we related categories to subcategorize along the lines of their properties and dimensions through the paradigm model. Finally, the grounded theory about the appropriation process was developed through the conditional/consequential matrix in selective coding. In this study eight hypotheses about the adaptation process have been clearly articulated. Also, we found that the appropriation process involves through three phases, namely, "direct appropriation," "cooperate with related structures," and "interpret and make judgments." The higher phases of appropriation move, the more users represent various types of instrumental use and attitude. Moreover, the previous structures like "knowledge and experience," "belief that other members know and accept the use of technology," "horizontal communication," and "embodiment of opinion collection process" are evolved to higher degrees in their dimensions of property. Furthermore, users continuously create new spirits and structures, while removing some of the previous ones at the same time. Thus, from longitudinal view, faithful and unfaithful appropriation methods appear recursively, but gradually faithful appropriation takes over the other. In other words, the concept of spirits and structures has been changed in the adaptation process over time for the purpose of alignment between the task and other structures. These findings call for a revised or extended model of structural adaptation in IS (Information Systems) literature now that the vague adaptation process in previous studies has been clarified through the in-depth qualitative study, identifying each phrase with accuracy. In addition, based on these results some guidelines can be set up to help determine which strategies are needed in an individual, group, and organizational level for the purpose of effective technology appropriation. In practice, managers can focus on the changes of spirits and elevation of the structural dimension to achieve effective technology use.

• Journal of Chest Surgery
• /
• v.34 no.7
• /
• pp.534-538
• /
• 2001

Background: Knowledge of size and morphology of the normal trachea is important for airway management and tracheal reconstruction. Conventional radiography is a simple method used to measure the tracheal diameter, but it is not accurate because of the artifacts related to image magnification and overlapping by the shoulder. The purpose of this study was to provide the normal values of the tracheal size and anatomy in Korean adults using Computerized Topography. Material and Method: There were 43 men and 34 women included in this study. They were divided into three age groups(group 1, 20-39 years ; group 2, 40-59 yeas , groups 3, $\geq$60 years). The anteroposterior and transverse diameters and cross - sectional areas of the trachea were measured at the level of the thoracic inlet(Level 1) and the aortic arch(Level 2). These values obtained at each level were compared between age groups and sexes. Result: In 43 men, the anteroposterior / transverse diameters(mean SD in millimeters) of the trachea at levels 1 and 2 were 19.95$\pm$2.99 / 17.72$\pm$2.13 and 19.77$\pm$2.57 / 18.02$\pm$2.19, respectively. In 34 women, those values at levels 1 and 2 were 15.56$\pm$2.12 / 14.18$\pm$2.07 and 15.35$\pm$1.82 / 15.00$\pm$1.60, respectively. At both levels, the anteroposterior and transverse diameters were significantly greater in men than in women (p<0.05). The cross-sectional area of the trachea at levels 1 and 2 were 279.14$\pm$61.37 / 281.93$\pm$63.97 $\textrm{mm}^2$ in men and 173.29$\pm$35.81 / 181.88$\pm$34.74 in women, respectively. They also showed significantly greater values in men than in women(P<0.05). There was no significant difference in diameters and cross-sectional areas of the trachea between age groups. Conclusion: There are significant differences in the internal diameter and cross- sectional area of the trachea between men and women in normal Korean adults, while the age difference was insignificant. We believed CT is a relatively accurate and safe way to measure the internal diameter and cross-sectional areas of the trachea.

• Journal of The Korean Association For Science Education
• /
• v.43 no.1
• /
• pp.49-57
• /
• 2023

The purpose of this study is to analyze the school biology curriculum of the Royal Society of Biology (RSB) for school biology education in the United Kingdom (UK), and to examine the implications for Korean school biology education. The RSB school biology curriculum consists of three dimensions (the practices of biology, the concepts of biology, and the application of biology) and seven big questions. The contents of RSB school biology curriculum are structured according to age, 5-11, 11-16, and 16-19. The practices of biology of the UK RSB school biology curriculum emphasize biology activities should be linked to other communities and communicated and shared through evidence-based discussions. The concepts of biology dimension of the UK RSB school biology curriculum systematized the contents to be detailed considering school-level and ages with 5-7, 7-11, 11-14, 14-16, and 16-19 years old. The UK RSB biology curriculum is composed of human health, biological structure and function, biological growth and reproduction, and biological coexistence, showing a significant difference from Korean national curriculum when comparing the content elements with the core concepts of biology. In terms of the application of biology of the RSB school biology curriculum, three themes were commonly applied to all students, regardless of school level or age, such as development of application methods to promote health and environmental welfare, evaluation of the impact and application of biology knowledge. However, the content elements of the three themes were systematized according to the ages of 5-11, 11-16, and 16-19 years old. It is necessary to reorganize the contents of biology curriculum in Korea by referring to the content system of the UK RSB biology curriculum and to research ways to strengthen connectivity according to age or school level as well as dimensions and big questions.

• Journal of Korean Home Economics Education Association
• /
• v.35 no.2
• /
• pp.145-161
• /
• 2023

The purpose of this study is to develop a content system for the home economics curriculum that integrates Education for Sustainable Development(ESD) and provides basic material for ESD implementation in schools. In view of this, the content elements of the revised home economics curriculum for 2022 were analyzed, and a content system for the home economics curriculum, linked to ESD, was proposed based on the implications drawn from the analysis. The results are as follows. First, the three components of competencies, namely knowledge, values, and skills, were organized equally as a whole. However, the association between the content elements and key competencies in sustainability was found to be insufficient. Consequently, it is proposed that key competencies in sustainability should be cultivated integrally. Second, no content element was identified that can promote social participation. Therefore, it is proposed that solutions should be dealt with at the level of social participation. Third, the connection with Sustainable Development Goals(SDGs) was observed in only six of the 28 content elements. Wherever relevant, it is proposed to incorporate key issues related to SDGs. Fourth, the analysis confirmed that only the environmental dimension of sustainable development was considered. Therefore, it is proposed to pursue coexistence based on temporal and spatial relationship and consider the dimensions of environment, society, and economy in an integrated manner.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.