• 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.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.7
• /
• pp.271-278
• /
• 2021

As the current cyber attacks become more intelligent, the existing Intrusion Detection System is difficult for detecting intelligent attacks that deviate from the existing stored patterns. In an attempt to solve this, a model of a deep learning-based intrusion detection system that analyzes the pattern of intelligent attacks through data learning has emerged. Intrusion detection systems are divided into host-based and network-based depending on the installation location. Unlike network-based intrusion detection systems, host-based intrusion detection systems have the disadvantage of having to observe the inside and outside of the system as a whole. However, it has the advantage of being able to detect intrusions that cannot be detected by a network-based intrusion detection system. Therefore, in this study, we conducted a study on a host-based intrusion detection system. In order to evaluate and improve the performance of the host-based intrusion detection system model, we used the host-based Leipzig Intrusion Detection-Data Set (LID-DS) published in 2018. In the performance evaluation of the model using that data set, in order to confirm the similarity of each data and reconstructed to identify whether it is normal data or abnormal data, 1D vector data is converted to 3D image data. Also, the deep learning model has the drawback of having to re-learn every time a new cyber attack method is seen. In other words, it is not efficient because it takes a long time to learn a large amount of data. To solve this problem, this paper proposes the Siamese Convolutional Neural Network (Siamese-CNN) to use the Few-Shot Learning method that shows excellent performance by learning the little amount of data. Siamese-CNN determines whether the attacks are of the same type by the similarity score of each sample of cyber attacks converted into images. The accuracy was calculated using Few-Shot Learning technique, and the performance of Vanilla Convolutional Neural Network (Vanilla-CNN) and Siamese-CNN was compared to confirm the performance of Siamese-CNN. As a result of measuring Accuracy, Precision, Recall and F1-Score index, it was confirmed that the recall of the Siamese-CNN model proposed in this study was increased by about 6% from the Vanilla-CNN model.

• Journal of the Korean earth science society
• /
• v.32 no.7
• /
• pp.860-870
• /
• 2011

This study is to explore the Pedagogical Content Knowledge of beginning earth science teachers about the porphyritic texture of igneous rocks, and to suggest the teaching device that can prevent a trial and error of students in earth science instruction. We developed an interview guideline concerned with basic perception on the porphyritic texture, formation condition and formation process of porphyritic rocks, teaching and learning on porphyritic rocks for it. And data was collected from 5 beginning earth science teachers (3 high schools, 2 middle schools) through a group discussion method. In result, despite the porphyritic texture can be found at hypabyssal rocks as well as volcano rocks and plutonic rocks, most beginning earth science teachers cognized that it could be found at hypabyssal rocks only by focusing the formation depth of hypabyssal rocks. Also, the formation of porphyritic texture should be considered the factors such as cooling rate, nucleation density, growth rate, growth time, etc. However they mainly reflected the formation temperature and growth rate as it's parameter. Participants have wrongly perceived that a phenocryst necessarily differs from a groundmass on chemical composition. And they are inclined to discriminate phenocryst from groundmass through their chemical differences, instead of grain size.

• Journal of The Korean Association For Science Education
• /
• v.40 no.2
• /
• pp.141-150
• /
• 2020

The purpose of this study is to critically review the skill-based approach to scientific inquiry in science education and to explore the meaning of science practices that are emphasized in recent science education reform movement. An extensive review of relevant literature was carried out, and the results were summarized according to the detailed themes of the study. In the skill-based approach of which Science-A Process Approach (SAPA) is a representative example, science process skills were presented as hierarchically connected with one another, they were believed to be transferable or generalizable, and science learning through discovery was stressed. These points of view are, however, contradicted with those of the modern philosophy of science which suggests the theory-laden nature of using the skills. The skill-based view has also been criticized by the fact that the use of inquiry skills is content-specific or context-dependent and that science theories or principles cannot be discovered by induction. In contrast, the recent view understands science practices holistically, emphasizes the diverse ways of doing the practices which vary with different contents or contexts, and considers student ideas importantly in the science classroom. The findings of this study can contribute to the development of a new science curriculum by providing implications for establishing a consistent view on scientific inquiry.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.8
• /
• pp.171-180
• /
• 2017

The purpose of this study is to determine an effective training method to improve sequela, since traumatic brain injury sequela is a major factor in determining the quality of life. Neurofeedback training was conducted for an adolescent who had experienced traumatic brain injury during his childhood and who had difficulty in cognitive learning and emotional aspects. The assessment of an adolescent was conducted using K-WAIS-IV intelligence test and QEEG brain wave analysis. In the neurofeedback training, T3 alpha wave compensation and T4 alpha wave inhibition training were performed 36 times for 30 minutes three times a week. In addition to the neurofeedback training, respiratory meditation was also made available to the adolescent. As a result, the adolescent showed a stable condition as indicated by taking a good sleep, reducing test anxiety, and satisfaction with final exam results. This study revealed the possibility for hidden physical and psychological problems arising due to childhood brain trauma. It has also recently been discovered that a more diverse set of tools can be found. In addition, these childhood traumatic brain injuries can be improved through brain training and meditation. The study finding is meaningful for its suggestion of a fusion method for developing mind and body therapy in terms of brain science.

• Journal of The Korean Association For Science Education
• /
• v.24 no.3
• /
• pp.565-582
• /
• 2004

The purpose of this study was to explore the elements of teaching behavior and classify instructional types through the teacher's classroom observation in elementary school science classrooms. 18 elementary school teachers were selected at Seoul city and Kyungkido. The topic of lesson was 'How the weight of object is changed according to the shape to sink in the water'. Each class was recorded and analyzed that. The teaching behavior elements were used inductional analysis method. The instruction types were classified into instructional organization, teaching strategies in teaching-learning processes, the level of openness of inquiry at science classroom. The validity and reliability of the data were analyzed by 7 science educators. The results of the analysis of the teachers discourse showed that there are 23 types of teaching behavior elements. Used teaching behavior elements revealed the differences from each teacher. There were 7 types among the 12 types of class and the most common types of instruction were unsystematic, teacher-centered, and guided-inquiry. The result showed that guided inquiry type was found more than open inquiry type and teacher-centered instructional, content-centered instructional, superficial inquiry process showed characteristic.

• Journal of The Korean Association For Science Education
• /
• v.37 no.3
• /
• pp.523-537
• /
• 2017

The purpose of this study is to investigate the features of elementary students' intuitive thinking emerged during problem solving activities as it related to thermal phenomena, focusing on when intuitive thinking appears and how it is related to logical thinking. For this, we presented a problem related to thermal phenomena to nine 5th-grade students, and examined how students' thinking emerged in the activities. We conducted clinical interviews to investigate the thinking process of students. The results of this study are as follows. First, students made their own solutions and justified it later during the emergence process of intuitive thinking. It was also found that students connected concrete materials and abstract concepts intuitively. They solved the problem by making predictions even when information is insufficient. Second, it was shown that intuitive thinking can emerge through the intended strategies such as drawing a mental image, thinking from a different perspective, and integrating methods. These results, which are related to the students' intuitive thinking has received little attention and will be the basis for helping students in the context of discovery of their problem solving activities.

• Journal of The Korean Association For Science Education
• /
• v.34 no.8
• /
• pp.755-765
• /
• 2014

The convergence and consilience in education (hereafter, interdisciplinary education) is receiving great attention from societies. This study aims to investigate the works of scientists and artists who have intended to combine science with the arts in the modern era, to take into account the socio-philosophical setbacks during the period, and to suggest pedagogical implications of science education as interdisciplinary education. The concept of interdisciplinary education stems from Plato's thought, idea, as a comprehensive and invariant truth. The renaissance, full of enrichment about scientific achievement, was based on Neo-Platonism pursuing holistic-synthetic approach. During the time, scientists presented in this study tried to find comprehensive principles and borrow useful method from the arts. In such a context, scientists not only made use of the arts for expression of scientific knowledge, but also drew conclusion by analogical reasoning between science and the arts. Artists, as well, relied upon anatomy and optics especially, to elaborate linear perspective and even developed their own scientific knowledge through personal experience. Hence, contemporary science education should encourage students to hold a holistic viewpoint about science and the arts, articulate explicit goals and outcomes as interdisciplinary education, implement meta-disciplinary instruction about science and the arts, and develop assessment framework for collaborative learning. There may be good examples for inter-disciplinary education as listed: illustrating scientific ideas through the arts and vice versa, organizing collaborative works and evaluations criteria for them, and stressing problem solving on a daily basis.

• Journal of the Korean School Mathematics Society
• /
• v.14 no.4
• /
• pp.423-442
• /
• 2011

The purpose of this study was to examine the effects of tasks setting for mathematical modelling in the complex real situations. The tasks setting(MMa, MeA) in mathematical modelling was so important that we can't ignore its effects to develop meaning and integrate mathematical ideas. The experimental setting were two groups ($N_1=103$, $N_2=103$) at public high school and non-experimental setting was one group($N_3=103$). In mathematical achievement, we found meaningful improvement for MeA group on modelling tasks, but no meaningful effect on information processing tasks. The statistical method used was ACONOVA analysis. Beside their achievement, we were much concerned about their modelling approach that TSG21 had suggested in Category "Educational & cognitive Midelling". Subjects who involved in experimental works showed very interesting approach as Exploration, analysis in some situation ${\Rightarrow}$ Math. questions ${\Rightarrow}$ Setting models ${\Rightarrow}$ Problem solution ${\Rightarrow}$ Extension, generalization, but MeA group spent a lot of time on step: Exploration, analysis and MMa group on step, Setting models. Both groups integrated actively many heuristics that schoenfeld defined. Specially, Drawing and Modified Simple Strategy were the most powerful on approach step 1,2,3. It was very encouraging that those experimental setting was improved positively more than the non-experimental setting on mathematical belief and interest. In our school system, teaching math. modelling could be a answer about what kind of educational action or environment we should provide for them. That is, mathematical learning.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.2
• /
• pp.15-21
• /
• 2009

Hearing loss is one of the most common birth defects among infants. Most of hearing-impaired children are not diagnosed until 1 to 3 years of age - which is too late for the critical period (6 month) for normal speech and language development. If a hearing impairment is identified and treated in its early stage, child's speech and language skills could be comparable to his or her normal-hearing peers. For these reasons, hearing screening at birth and throughout childhood is extremely important. ABR (Auditory brain-stem response) is nowadays one of the most reliable diagnostic tools in the early detection of hearing impairment. In this study, we have developed the system that automatically detects if there is hearing impairment or not for infants or children. For future studies, it will be developed as a portable system to be able to take a measurement not only in sound proof room but also in nursery for neonates.