• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.4
• /
• pp.819-833
• /
• 2013

Computing global features based on local features using a bio-inspired framework has shown promising performance. However, for some tough applications with large intra-class variances, a single local feature is inadequate to represent all the attributes of the images. To integrate the complementary abilities of multiple local features, in this paper we have extended the efficacy of the bio-inspired framework, HMAX, to adapt heterogeneous features for global feature extraction. Given multiple global features, we propose an approach, designated as parameterized metric learning, for high dimensional feature fusion. The fusion parameters are solved by maximizing the canonical correlation with respect to the parameters. Experimental results show that our method achieves significant improvements over the benchmark bio-inspired framework, HMAX, and other related methods on the Caltech dataset, under varying numbers of training samples and feature elements.

• School Mathematics
• /
• v.1 no.1
• /
• pp.19-37
• /
• 1999

In this study, we examined some matters related to the learning of variable concept in school mathematics on the basis of the theoretical foundation from the previous studies(e.g. Davis, 1975; Rosnick, 1981; IK chemann, 1981; Wagner, 1983.) and practices on variable concept teaching by evaluating the current state of that. Matters be discussed are as follows; the use of symbol for place holders in elementary mathematics, the dealing with sets those elements are literals and operations of such sets, the teaching of dummy variable, the construction of literal expressions that contains variables, the labeling indeterminates as a constant, the change in the exact meaning of variable according to the function concept, the teaching of a generalization by means of variables. After considering on these matters that are connected with the teaching-learning of variable concept, we suggested the alternative proposal to the current state of variable concept teaching.

• School Mathematics
• /
• v.1 no.1
• /
• pp.109-122
• /
• 1999

In this paper we give a description of students' obstacles in their learning of geometry, especially resulted from their confusing a physical drawing with a figure, a geometrical object which a physical drawing represents. In computer-based teaching-learning environment, we could relieve such obstacles through providing students for experiences in which they must focus on elements of a figure and relations of them. But there may be potential in computer-based environment if we offer students only visual experience for validity of geometrical gacts: students' lack of understanding for need of proof and experience of cognitive obstacles which is very important for students to reflect their thinking and activities. Thus an didactical treatment must follows, which we also give an example.

• Journal of Korean Library and Information Science Society
• /
• v.39 no.2
• /
• pp.449-467
• /
• 2008

In this study, collaboration of teacher librarians and subject teachers in school libraries has been investigated(tried, attempted, suggested) in order to improve the information literacy. The effective teaching-learning can be conducted by extending student's roles, which are essential elements in school education. Above all, the model. focused on direct exchange of knowledges in communications between teachers and students has been suggested with consideration of content knowledge, pedagogical knowledge, and technological knowledge.

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10a
• /
• pp.830-835
• /
• 1993

The errors in machining process by CNC machining center are due to many elements, such as the delay of the servo drivers, friction and the gain mismatch between x-axis and y-axis motors and so on. We made a counter circuit to measure the output of motor encoders for the motion error analysis of a CNC machining center, and have measured the errors experimentally when the CNC performs a circular interpolation. We have also used an iterative learning method to reduce the radius errors and stick motion errors generated by the CNC machining center performing a circular interpolation. The proposed learning scheme worked well and the circle obtained has smaller error.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1996.10a
• /
• pp.243-248
• /
• 1996

In this paper, a general neural-network-based connectionist model, called Fuzzy Neural Network(FNN), is proposed for the realization of a fuzzy logic control system. The proposed FNN is a feedforward multi-layered network which integrates the basic elements and functions of a traditional fuzzy logic controller into a connectionist structure which has distributed learning abilities. Such FNN can be constructed from training examples by learning rule, and the connectionist structure can be trained to develop fuzzy logic rules and find optimal input/output membership functions. Computer simulation examples will be presented to illustrate the performance and applicability of the proposed FNN, and their associated learning algorithms.

• Proceedings of the KIEE Conference
• /
• 1998.07g
• /
• pp.2296-2298
• /
• 1998

In this paper, an on-chip learning pulse-mode digital neural network with a massively parallel yet compact and flexible network architecture is suggested. Algebraic neural operations are replaced by stochastic processes using pseudo-random sequences and simple logic gates are used as basic computing elements. Using Back-propagation algorithm both feed-forward and learning phases are efficiently implemented with simple logical gates. RNG architecture using LFSR and barrel shifter are adopted to avoid some correlation between pulse trains. Suggested network is designed in digital circuit and its performance is verified by computer simulation.

• Journal of Fisheries and Marine Sciences Education
• /
• v.17 no.3
• /
• pp.373-382
• /
• 2005

The aim of this study is to analyze critically the meaning of 'good instruction'. In this article, the researchers emphasized the activities of teaching and learning as two essential elements of good instruction. The antecedent studies on good instruction have limitations in that they regarded good instruction as an instrument rather than valuable activity itself. This study consisted of three parts. Firstly, the researchers tried to clarify the meaning of good instruction underlying antecedent studies on instruction. Secondly, the researchers tried to analyze characteristics of good insturction suggested by antecedent studies. Thirdly, the researchers attempted to define a new meaning of good instruction. A new meaning of good instruction should be defined in accordance with characteristics of three commonplaces composing instruction such as teaching activity, learning activity and good interactions between teacher and learner. Good instruction can be defined as followings. Good instruction means an instructional activity which can change the student's meanings of experience as a result of interactions among teacher, learner, and educational materials.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.8
• /
• pp.578-584
• /
• 2004

The CSVM(Current Support Vector Machine) that is a digital architecture performing all phases of recognition process including kernel computing, learning, and recall of SVM(Support Vector Machine) on a chip is proposed. Concurrent operation by parallel architecture of elements generates high speed and throughput. The classification problems of bio data having high dimension are solved fast and easily using the CSVM. Quadratic programming in original SVM learning algorithm is not suitable for hardware implementation, due to its complexity and large memory consumption. Hardware-friendly SVM learning algorithms, kernel adatron and kernel perceptron, are embedded on a chip. Experiments on fixed-point algorithm having quantization error are performed and their results are compared with floating-point algorithm. CSVM implemented on FPGA chip generates fast and accurate results on high dimensional cancer data.

• Journal of Korean language education
• /
• v.23 no.1
• /
• pp.25-45
• /
• 2012

The purpose of this article is to examine types and construction method of multimedia materials for the Korean language education, finally in order to construct digital library on Nuri-Sejonghakdang. Firstly, this article reviews some major concepts such as teaching material, multimedia, learning object, meta-data, reusability, etc. Secondly, various multimedia materials are divided into three types(namely, example material, explanation material, training and evaluating material) according to their characteristics as a learning objects. And then, this article tries to propose the classification-search system and meta-data elements for effective search and use of multimedia materials. Finally, this article is concluded by presenting the long-term plan of digital library construction on Nuri-Sejonghakdang and some follow-up task of this study.