• Journal of Sensor Science and Technology
• /
• v.10 no.6
• /
• pp.286-294
• /
• 2001

In this study, 3-D underwater object recognition using the self-made 3-3 type composite ultrasonic transducer and modified SOFM(Self Organizing Feature Map) neural network are investigated. Properties of the self-made 3-3 type composite specimens are satisfied considerably with requirements as an underwater ultrasonic transducer's materials. 3-D underwater all object's recognition rates obtained from both the training data and testing data in different objects, such as a rectangular block, regular triangular block, square block and cylinderical block, were 100% and 94.0%, respectively. All object's recognition rates are obtained by utilizing the self-made 3-3 type composite transducer and SOFM neural network. From the object recognition rates, it could be seen that an ultrasonic transducer fabricated with the self-made 3-3 type composite resonator will be able to have application for the underwater object recognition.

• Journal of Embryo Transfer
• /
• v.12 no.3
• /
• pp.283-291
• /
• 1997

In vitro development of bovine embryos is affected by many factors such as energy substrates, amino acids, and some growth factors. It has been reported that mRNA of insulin, PDGF and their receptors are detected in cow embryos, and that some chelating agents such as EDTA and transferrin have beneficial role on mouse and bovine embryos. The author hypothesized that insulin, transferrin arid PDGF added to a culture medium increase in vitro development of bovine embryos by chelating toxic substance(s) or increasing cell growth and metabolism. Immature oocytes from slaughtered ovaries of Holstein cows and heifers were matured for 24 hours in a TCM199 containing 10% fetal calf serum, FSH, LH and estradiol with granulosa cells in vitro. Matured oocytes were coincubated with sperm for 30 hours in a modified Tyrode's medium (IVF). Embryos cleaved to 2- to 4-cell at 30 hours after IVF were selected and cultured in a 30-$\mu$l drop of a synthetic oviduct fluid medium (SOFM) containing 0.8% BSA, Minimum Essential Medium essential and non-essential amino acids, and insulin, transferrin or PDGF for 9 days. Supplementation of a SOFM with insulin, and /or transferrin did not increase develop-mental rate to expanding and hatching blastocyst of 2- to 4-cell bovine embryos compared with control. The highest developmental rate to hatching blastocyst was shown when PDGF was added at the concentration of 10 ng /ml among the supplementing doses tested in the present study (p<0.05). Addition of PDGF without insulin to a SOFM could not increase embrye development, but combined addition of PDGF with insulin significantly increased (p<0.05) embryo development to hatching blastocyst (50%) compared with control (38%). In conclusion, insulin and PDGF supplemented to a SOFM may act synergistically and have beneficial effect on in vitro development of 2- to 4-cell bovine embryos matured and fertilized in vitro.

• Journal of Practical Engineering Education
• /
• v.6 no.2
• /
• pp.71-76
• /
• 2014

Improving lecture quality is very necessary for the service quality of education in universities, enterprises and education institutes. The student lecture evaluation survey data is a good tool for measuring lecture quality and have been often analyzed by simple statistical methods. This study presents an intelligent lecture quality improvement method that can improve student's overall satisfaction and performance by analyzing student lecture evaluation survey data. The method uses SOFM (Self-Organizing Feature Map) neural network and C4.5 to find the patterns in student's satisfaction and performance more correctly and then decide what to change in the lecture for the improvement of student's satisfaction and performance. We apply the proposed method to an enterprise lecture in Korea. We can find that it can improve the quality of an enterprise lecture by changing total lecture time, lecture material and organization of lecture schedule to be necessary improvements.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1039-1043
• /
• 2005

Despite the many significant advances made in robot architecture, the basic approaches are deliberative and reactive methods. They are quite different in recognizing outer environment and inner operating mechanism. For this reason, they have almost opposite characteristics. Later, researchers integrate these two approaches into hybrid architecture. In such architecture, Reactive module also called low-level motion control module have advantage in real-time reacting and sensing outer environment; Deliberative module also called high-level task planning module is good at planning task using world knowledge, reasoning and intelligent computing. This paper presents a framework of the integrated planning and control for mobile robot navigation. Unlike the existing hybrid architecture, it learns topological map from the world map by using MST (Minimum Spanning Tree)-based SOFM (Self-Organizing Feature Map) algorithm. High-level planning module plans simple tasks to low-level control module and low-level control module feedbacks the environment information to high-level planning module. This method allows for a tight integration between high-level and low-level modules, which provide real-time performance and strong adaptability and reactivity to outer environment and its unforeseen changes. This proposed framework is verified by simulation.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.5
• /
• pp.85-91
• /
• 1989

In this paper, we propose a neural net system for speech recognition, which is composed of two neural networks. Firstly the self-supervised BP(Back Propagation) network generates the distributed concept corresponding to the activity pattern in the hidden units. And then the self-organizing neural network forms a concept map which directly displays the similarity relations between concepts. By doing the above, the difficulty in learning the conventional BP network is solved and the weak side of BP falling into a pattern matcher is gone, while the strong point of generating the various internal representations is used. And we have obtained the concept map which is more orderly than the Kohonen's SOFM. The proposed neural net system needs not any special preprocessing and has a self-learning ability.

• Journal of Internet Computing and Services
• /
• v.11 no.4
• /
• pp.85-94
• /
• 2010

In this paper, we propose a location-prediction structure that can provide user service in advance. It consists of seven steps and supplies intelligent services which can forecast user's location. Context information collected from physical sensors and a history database is so difficult that it can't present importance of data and abstraction of data because of heterogeneous data type. Hence, we offer the location-prediction that change data type from heterogeneous data to homogeneous data. Extracted data is clustered by SOFM, then it gets user's location information by ARIMA and realizes the services by a reasoning engine. In order to validate the proposed location-prediction, we built a test-bed and test it by the scenario.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.449-451
• /
• 1999

지금까지 많은 필기 숫자 인식 방법들이 제안되었지만 고도의 신뢰도가 요구되는 은행 수납 장표상의 숫자 인식에 적합한 방법은 아직 발표된 것이 미미한 실정이다. 본 연구에서는 세 개의 분류기의 결과를 융합하여 100%에 가까운 신뢰도를 낼 수 있는 필기숫자 인식 시스템을 제안하였다. Karhunen-Loeve Transform(KLT)를 통하여 특징을 추출하였으며 오류 역전파 신경망(BP), LVQ를 적용한 SOFM(SOFM-LVQ)과 Weignted Several Nearest Neighbor(WSNN)을 분류기로 사용하였다. 융합을 위해서는 다수결(Majority Voting)이 아닌 만장일치제(Unanimous Voting)을 적용하여 신뢰도를 높혔다. ETL-6 DB를 사용하여 실험하였으며 실험 결과 99.95%의 높은 신뢰도를 기록하였다.

• Journal of Institute of Control, Robotics and Systems
• /
• v.2 no.2
• /
• pp.108-114
• /
• 1996

We present proofs of the stability and convergence of Self-organizing feature map (SOFM) neural network with time-invarient learning rate and binary reinforcement function. One of the major problems in Self-organizing feature map neural network concerns with learning rate-"Kalman Filter" gain in stochsatic control field which is monotone decreasing function and converges to 0 for satisfying minimum variance property. In this paper, we show that the stability and convergence of Self-organizing feature map neural network with time-invariant learning rate. The analysis of the proposed algorithm shows that the stability and convergence is guranteed with exponentially stable and weak convergence properties as well.s as well.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.10a
• /
• pp.402-408
• /
• 1997

일정 학습계수와 이진 강화함수를 지닌 자기조직화 형상지도(Self-Organizing Feature Map)신경회로망을 FPGA위에 하드웨어로 구현하였다. 원래의 SOFM 알고리즘에서 학습계수가 시간 종속형인데 반하여, 본 논문에서 하드웨어로 구현한 알고리즘에서는 학습계수가 일정인 값으로 고정되며 이로 인한 성능저하를 보상하기 위하여 이진 강화함수를 부가하였다. 제안한 알고리즘은 복잡한 곱셈 연산을 필요로 하지 않으므로 하드웨어 구현시 보다 쉽게 구현 가능한 특징이 있다. 1개의 덧셈/뺄셈기와 2개의 덧셈기로 구성된 단위 뉴런은 형대가 단순하면서 반복적이므로 하나의 FPGA위에서도 다수의 뉴런을 구현 할 수 있으며 비교적 소수의 제어 신호로서 이들을 모두 제어 가능할 수 있도록 설계하였다. 실험결과 각 구성부분은 모두 이상 없이 올바로 동작하였으며 각 부분이 모두 종합된 전체 시스템도 이상 없이 동작함을 알 수 있었다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.10a
• /
• pp.1032-1035
• /
• 1997

The purpose of this paper is to reduce the number of cutting tests and to predict the main cutting force and the specific cutting energy. By using the SOFM neural network, the most suitable cutting test conditions has been found. As a result, the number of cutting tests has been reduced to one-third. And by using MLP neural network and regression analysis, the main cutting force and specific cutting energy has been predicted. Predicted values of main cutting force and specific cutting energy are well concide with the measured ones.