• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.49 no.2
• /
• pp.115-122
• /
• 2012

We present here a new class of activation functions for neural networks, which herein will be called CosGauss function. This function is a cosine-modulated gaussian function. In contrast to the sigmoidal-, hyperbolic tangent- and gaussian activation functions, more ridges can be obtained by the CosGauss function. It will be proven that this function can be used to aproximate polynomials and step functions. The CosGauss function was tested with a Cascade-Correlation-Network of the multilayer structure on the Tic-Tac-Toe game and iris plants problems, and results are compared with those obtained with other activation functions.

• Korean Journal of Optics and Photonics
• /
• v.24 no.2
• /
• pp.64-70
• /
• 2013

In this paper, two white light emitting diodes(LEDs) were manufactured by using two kinds of yellow phosphor, YAG:Ce and $(Sr,Ba)_2SiO_4:Eu$, and their spectroscopic properties were compared and analyzed. We found that the asymmetric double sigmoidal function can be applied to both white LEDs to obtain the center wavelength, the half width, and the skewness parameters. According to this analysis, the half width of the emitting spectrum of silicate phosphor was smaller than that of YAG phosphor, indicating smaller color rendering index. However, the silicate phosphor exhibited better color stability depending on the driving current than the YAG phosphor. The current dependence of the luminous efficacy of both white LEDs was investigated, which showed that efficacy decreased monotonically with current. The efficacy of the silicate-based white LED was lower than that of the YAG-based LED by about 10~12 lm/W.

• International Journal of Computer Science & Network Security
• /
• v.21 no.2
• /
• pp.77-87
• /
• 2021

In this study, a novel improved second order Radial Basis Function Neural Network based method with excellent scheduling capabilities is used for the dynamic prediction of short and long-term energy required applications. The effectiveness and the reliability of the algorithm are evaluated using training operations with New England-ISO database. The dynamic prediction algorithm is implemented in Matlab and the computation of mean absolute error and mean absolute percent error, and training time for the forecasted load, are determined. The results show the impact of temperature and other input parameters on the accuracy of solar Photovoltaic load forecasting. The mean absolute percent error is found to be between 1% to 3% and the training time is evaluated from 3s to 10s. The results are also compared with the previous studies, which show that this new method predicts short and long-term load better than sigmoidal neural network and bagged regression trees. The forecasted energy is found to be the nearest to the correct values as given by England ISO database, which shows that the method can be used reliably for short and long-term load forecasting of any electrical system.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.1
• /
• pp.30-35
• /
• 2010

This paper presents a quality enhancement of images by using a histogram equalization based on the symmetric logistic function. The histogram equalization is a simple and effective spatial processing method that it enhances the quality by adjusting the brightness of image. The logistic function that is a sigmoidal nonlinear transformation function, is applied to non-linearly enhance the brightness of the image according to its intensity level frequency. We propose a flexible and symmetrical logistic function by only using the intensity with maximum frequency in an histogram and the total number of pixels. The proposed function decreases the computation load of an exponential function in the traditional logistic function. The proposed method has been applied for equalizing 5 images with a different resolution and histogram distribution. The experimental results show that the proposed method has the superior enhancement performances compared with the source images and the traditional global histogram equalization, respectively.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2003.11a
• /
• pp.240-240
• /
• 2003

탄소재료의 가스화속도는 근본적으로 활성자리의 수와 관련되어 있으며, 또한 가스화속도는 활성자리 뿐 아니라 확산제한에 따라 달라진다. 대부분의 탄소재료의 활성화 초기단계는 제한된 활성자리 때문에 반응속도는 느리고, 다음 단계는 총 활성자리가 증가하여 반응속도는 급격히 증가하고, 마지막으로 활성자리가 감소하여 활성화 속도는 감소한다. 이러한 sigmoidal특성을 나타내는 활성화 단계를 기공발달과정으로 설명하면, 활성화 초기에 탄소재료 내부에 이미 존재하는 닫힌 기공이 열리고, 일단 기공이 열리면 성장하게 된다. 이렇게 기공 수가 증가하는 것 뿐 아니라 기공 직경이 증가하여 활성화 과정이 진행될수록 비 표면적 및 기공부피는 증가하는데 이런 일련의 과정을 통하여 활성자리 수는 증가하고 또는 감소한다. 이렇게 기공이 발달하는 과정은 각각의 활성화 단계에서 탄소재료의 비 표면적 측정으로 알 수 있으며, 전반적인 산화속도 변화를 측정하여 반응단계를 추정하게 된다. 대부분의 연구자들은 반응 전체의 평균 산화속도를 측정한 후 활성화 에너지를 구하여 반응조절단계로 활성화 기구를 설명한다. 이 연구에서는 활성화 과정 중에 발생하는 중량감소 단계, 즉 각각의 활성화 단계에 따라 달라지는 반응속도상수를 측정하고, 반응단계별 활성화 에너지를 비교 해석하여 피치계 탄소섬유의 기공발달에 영향을 미치는 활성화 기구를 고찰하고자 하였다.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.31A no.8
• /
• pp.62-71
• /
• 1994

A novel semiconductor device is proposed to be used as a processing element for the implementation of pulse-mode neural networks which consists of alternating n' GaAs quantum wells and undoped AlGaAs barriers sandwitched between n' GaAs cathode and P' GaAs anode and in simple circuit in conjunction with a parallel capacitive and resistive load the trigger circuit generates neuron-like pulse train output mimicking the function of axon hillock of biological neuron. It showed the sigmoidal relationship between the frequency of the pulse-train and the applied input DC voltage. In conjunction with MQWIMD the various neural circuits are proposed especially a neural chip monolithically integrated with photodetectors in order to perfrom the pattern recognition.

• Bulletin of the Korean Chemical Society
• /
• v.5 no.4
• /
• pp.154-158
• /
• 1984

In order to characterize the permeability of small unilamellar vesicles (SUV), efflux of 6-carboxyfluorescein (6-CF) from the vesicles was monitored spectrophotofluorometrically. Since the entrapped highly quenched 6-CF (200 mM) became fluorescent upon release from the vesicles, the 6-CF could be used as an efflux probe. SUV containing entrapped 6-CF was prepared from egg phosphatidylcholine and separated by gel filtration on Sepharose 4B. Observed change of relative fluorescent intensity with time was sigmoidal. From this curve, the parameter of permeability was determined either by half-time or a released amount per unit time from the initial slope. Half-time of efflux of prepared SUV having 302 ng phospholipid/ml in 10 mM Tris-HCl buffer pH 7.4 was 21.0 min at $37{\circ}C$. Various factors which could affect the half-time were examined including temperature, pH, salt, and vesicle concentration. In particular the effect of vesicle concentration on the efflux revealed that the permeability can be a function of the concentration.

• The Journal of the Korea Contents Association
• /
• v.10 no.4
• /
• pp.10-18
• /
• 2010

This paper presents a optimization of sigmoid activation function parameter using genetic algorithms and pattern recognition analysis in input space of two spirals benchmark problem. To experiment, cascade correlation learning algorithm is used. In the first experiment, normal sigmoid activation function is used to analyze the pattern classification in input space of the two spirals problem. In the second experiment, sigmoid activation functions using different fixed values of the parameters are composed of 8 pools. In the third experiment, displacement of the sigmoid function to determine the value of the three parameters is obtained using genetic algorithms. The parameter values applied to the sigmoid activation functions for candidate neurons are used. To evaluate the performance of these algorithms, each step of the training input pattern classification shows the shape of the two spirals.

• Journal of Power Electronics
• /
• v.19 no.4
• /
• pp.835-845
• /
• 2019

This paper presents a robust-optimal control strategy to improve the output-voltage error-tracking and control capability of a DC-DC boost converter. The proposed strategy employs an optimized Fractional-order Proportional-Integral (FoPI) controller that serves to eliminate oscillations, overshoots, undershoots and steady-state fluctuations. In order to significantly improve the error convergence-rate during a transient response, the FoPI controller is augmented with a pre-stage nonlinear error-modulator. The modulator combines the variations in the error and error-derivative via the signed-distance method. Then it feeds the aggregated-signal to a smooth sigmoidal control surface constituting an optimized hyperbolic secant function. The error-derivative is evaluated by measuring the output-capacitor current in order to compensate the hysteresis effect rendered by the parasitic impedances. The resulting modulated-signal is fed to the FoPI controller. The fixed controller parameters are meta-heuristically selected via a Particle-Swarm-Optimization (PSO) algorithm. The proposed control scheme exhibits rapid transits with improved damping in its response which aids in efficiently rejecting external disturbances such as load-transients and input-fluctuations. The superior robustness and time-optimality of the proposed control strategy is validated via experimental results.

• The KIPS Transactions:PartD
• /
• v.11D no.3
• /
• pp.597-604
• /
• 2004

Accurate software cost estimation is essential to both developers and customers. Most of the cost estimating models based on the size measure methods, such as LOC and FP, are obtained through size estimation. The accuracy of size estimation directly influences the accuracy of cost estimation. As a result, the overall structure of regression-based cost models applies the power function based on software size. Many growth phenomenon in nature such as the growth in living organism, performance of technology, and learning capability of human show an S-shaped curve. This paper proposes a model which estimates the developing effort by using the growth curve. The presented model assumes that the relation cost and size follows the growth curve. The appropriateness of the growth curve model based on Function Point, Full-Function Point and Use-Case Point, which are the general methods in estimating the software size have been confirmed. The proposed growth curve model shows similar performance with power function model. In conclusion, the growth curve model can be applied in the estimation of the software cost.