• Journal of KIISE:Software and Applications
• /
• v.28 no.9
• /
• pp.681-687
• /
• 2001

In order to raise a class discrimination power by combining multiple classifiers under the Bayesian decision theory, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables obtained from training data samples should be minimized. Wang and Wong proposed a tree dependence first-order approximation scheme of a high order probability distribution composed of the class and multiple feature pattern variables for minimizing the upper bound of the Bayes error rate. This paper presents an extended high order product approximation scheme dealing with higher order dependency more than the first-order tree dependence, based on the minimization of the upper bound of the Bayes error rate. Multiple recognizers for unconstrained handwritten numerals from CENPARMI were combined by the proposed approximation scheme using the Bayesian formalism, and the high recognition rates were obtained by them.

• Journal of the Korean School Mathematics Society
• /
• v.3 no.1
• /
• pp.149-163
• /
• 2000

The Ministry of Education have had us practice the performance test as a substitute proposal, however, all the more for the idealistic purport, our education front does not have such a sufficient condition as to practice the performance test for many classes and miscellaneous duties and over-populated class, and that practice has been enforced so abruptly without any drastic preparation and has caused much confusion from the beginning of that enforcement. Thus, these problematic concerns are remained as the tasks of the teachers to be solved by themselves in the front of education, and herein I came to do this research. The followings are the conclusions that I got as the results of the research (1) Performance test style should be applied in consideration of the students' achievement level and the gap of the teachers' recognition; descriptive test, portfolio assignment and formative test styles were proper for the students lacking basic study ability. (2) Descriptive test should have its beginning with the question items to which students can write the problem solving procedure logically rather than those to evaluate the creation ability and thinking ability: and putting down specifically the assessment standard could prevent students' confusion and scheme the impartiality of the assessment. (3) Portfolio assignment evaluation should be given with as interesting and suitable amounts as possible so that the students can do by themselves. (4) Utilizing the performance test table enabled easy management of documentary evidence. And it is needless to say that the success of the performance test should have preceding conditions like the teachers' understanding and their positive participation. Therefore, I'd like to give suggestions herein like the followings; (1) The performance test should not always be made into grades, and there is a need to develop the test gradually in the condition that the education surroundings permit by checking time, frequency, ratio and contents of the test while practicing the multiple choice writing test. (2) As long as the performance test has the aims of improving the studying and learning activities, any performance test only for the sake of making numerals with the thought that assessment is the disposal of the grades should be avoided, and the change of the lecturing styles and development of various assessing types and studying materials should be endeavored to confirm with the aims.

• Communications of Mathematical Education
• /
• v.28 no.3
• /
• pp.375-394
• /
• 2014

The purpose of this study is to examine the transition of mathematical representation in Joseon mathematics, which is focused on numbers and operations, letters and expressions. In Joseon mathematics, there had been two numeral systems, one by chinese character and the other by counting rods. These systems were changed into the decimal notation which used Indian-Arabic numerals in the late 19th century passing the stage of positional notation by Chinese character. The transition of the representation of operation and expressions was analogous to that of representation of numbers. In particular, Joseon mathematics represented the polynomials and equations by denoting the coefficients with counting rods. But the representation of European algebra was introduced in late Joseon Dynasty passing the transitional representation which used Chinese character. In conclusion, Joseon mathematics had the indigenous representation of numbers and mathematical symbols on our own. The transitional representation was found before the acceptance of European mathematical representations.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.8
• /
• pp.12-18
• /
• 2001

In this paper, we propose the Recurrent Neural Predictive HMM (RNPHMM). The RNPHMM is the hybrid network of the recurrent neural network and HMM. The predictive recurrent neural network trained to predict the future vector based on several last feature vectors, and defined every state of HMM. This method uses the prediction value from the predictive recurrent neural network, which is dynamically changing due to the effects of the previous feature vectors instead of the stable average vectors. The models of the RNPHMM are Elman network prediction HMM and Jordan network prediction HMM. In the experiment, we compared the recognition abilities of the RNPHMM as we increased the state number, prediction order, and number of hidden nodes for the isolated digits. As a result of the experiments, Elman network prediction HMM and Jordan network prediction HMM have good recognition ability as 98.5% for test data, respectively.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.40 no.5
• /
• pp.266-279
• /
• 2003

In this paper, we propose machine printed character recognition methods which utilize the character type information and divide the character clusters. The characters are subdivided into a total of seven types, of which six types are for Hangul according to the grapheme combination fashions and one type for English characters, numerals, and symbols. According to the character type, we separate input character image into several recognition units and recognize them by using the direction angle feature. The recognition for each character type is completed by combining recognition units which are recognized by neural networks respectively For combining a total of seven character recognizers, we implemented seven methods such as switching method, integrating method, and their several variants. As experimental results, we obtained 98.2% recognition rate of simple switching method, 90.54% of integrating one, and between 97.35% and 98.65% of five variants.

• Journal for History of Mathematics
• /
• v.21 no.3
• /
• pp.97-112
• /
• 2008

In this study, the research results up to now on original word form and its meaning of Korean number words hana, dul, ..., yeol are looked out from math-historically. In fact, finding out original word form and its meaning of hana, dul, and set(ses) may not be possible in the respect of history of mathematics. There might have been a gap between set(ses) and net(nes), and between net(nes) and daseot(daseos). Original word form and its meaning of hana, dul, set(ses), and net(nes) must be found out in different aspect from those of daseot(daseos), yeoseot(yeoseos), ..., yeol. There might have been a gap between yeoseot(yeoseos) and ilgop(ilgob). Coining number word mechanism for ilgop(ilgob), yeodeol,(yeodeolb) and ahop(ahob) might have been same each other. There might have been a gap between ahop(ahob) and yeol. The research results up to now have not paid attention to this gaps sufficiently. But according to history of mathematics, there must have existed several gaps.

• Korean Journal of Cognitive Science
• /
• v.7 no.1
• /
• pp.75-84
• /
• 1996

In this paper we emphasize the importance of architectural aspects of designing a handwritten numeral recognition program. and describe two architectural design.First, we describe the modular design of a numeral recognition program, and mention its advantages.In this design, a recognizer is composed of 10 binary subrecognizers each of which is responsible for only one class.Rule-based training and neural-based training are presented.Second, we connect two(or more)recognizers serially which we call pipelining connection.The second recognizer may act as verifier for the patterns recognized by the forst recognizer, or as second chance recognizer for the patterns rejected by the first recognizer.Our experimental results obtained till now show the merits of the proposed architectural designs.

• The Journal of the Korea Contents Association
• /
• v.13 no.10
• /
• pp.19-26
• /
• 2013

In the Malaysian license plates, alphabets and numerals which are made by plastic, are adhered to a frame as embossing style and occasionally characters in horizontal, vertical directions are aligned with narrow space. So the extraction of character stroke information can be hard in the vehicle images of low illumination intensity. In this paper, Malaysian license plate recognition algorithm for low illumination intensity image is proposed. DoG filtering based character stroke generation method is introduced to derive exact connected components of strokes in the vehicle image of low illumination intensity. After localization of plate by connected component analysis, characters are segmented and recognized. Algorithm is experimented for the 6,046 vehicle images captured in Kuala Lumpur by IR camera without using any special light during day and night. The experimental results show that recognition accuracy of plates is 96.1%.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.1
• /
• pp.44-52
• /
• 2001

This paper describes an off-line method for recognizing totally unconstrained handwritten digits using modified chaotic neural networks(MCNN). The chaotic neural networks(CNN) is modified to be a useful network for solving complex pattern problems by enforcing dynamic characteristics and learning process. Since the MCNN has the characteristics of highly nonlinear dynamics in structure and neuron itself, it can be an appropriate network for the robust classification of complex handwritten digits. Digit identification starts with extraction of features from the raw digit images and then recognizes digits using the MCNN based classifier. The performance of the MCNN classifier is evaluated on the numeral database of Concordia University, Montreal, Canada. For the relative comparison of recognition performance, the MCNN classifier is compared with the recurrent neural networks(RNN) classifier. Experimental results show that the classification rate is 98.0%. It indicates that the MCNN classifier outperforms the RNN classifier as well as other classifiers that have been reported on the same database.

• Communications of Mathematical Education
• /
• v.31 no.1
• /
• pp.149-177
• /
• 2017

Since modern mathematics textbooks were introduced in the late 19th century Korea, arithmetic experts started to teach modern mathematics using Arabic numerals at village schools and churches. After the Gabo Education Reform of 1894, western mathematics education was included in public education and the mathematics textbooks began to be officially published. We explored most of Korean mathematics textbooks from 1895 to 2016 including the changes of mathematics curriculum through 1885-1905, 1905-1910, 1911-1945, 1945-1948, 1948-1953, 1954-1999, and 2000-2016. This study presents the characters of modern mathematics textbooks of Korea since 1885.