• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.6
• /
• pp.594-601
• /
• 1997

A new algorithm is proposed for training the multilayer perceptrion(MLP) in pattern classification problems to accelerate the learning speed. It is shown that the sigmoid activation function of the output node can have deterimental effect on the performance of learning. To overcome this detrimental effect and to use the information fully in supervised learning, an objective function for binary modes is proposed. This objective function is composed with two new output activation functions which are selectively used depending on desired values of training patterns. The effect of the objective function is analyzed and a training algorithm is proposed based on this. Its performance is tested in several examples. Simulation results show that the performance of the proposed method is better than that of the conventional error back propagation (EBP) method.

• ETRI Journal
• /
• v.17 no.1
• /
• pp.11-22
• /
• 1995

This paper proposes a modified error function to improve the error back-propagation (EBP) algorithm for multi-Layer perceptrons (MLPs) which suffers from slow learning speed. It can also suppress over-specialization for training patterns that occurs in an algorithm based on a cross-entropy cost function which markedly reduces learning time. In the similar way as the cross-entropy function, our new function accelerates the learning speed of the EBP algorithm by allowing the output node of the MLP to generate a strong error signal when the output node is far from the desired value. Moreover, it prevents the overspecialization of learning for training patterns by letting the output node, whose value is close to the desired value, generate a weak error signal. In a simulation study to classify handwritten digits in the CEDAR [1] database, the proposed method attained 100% correct classification for the training patterns after only 50 sweeps of learning, while the original EBP attained only 98.8% after 500 sweeps. Also, our method shows mean-squared error of 0.627 for the test patterns, which is superior to the error 0.667 in the cross-entropy method. These results demonstrate that our new method excels others in learning speed as well as in generalization.

• Korean Journal of Adult Nursing
• /
• v.29 no.2
• /
• pp.190-199
• /
• 2017

Purpose: The purpose of this study was to investigate the differences in sleep, fatigue, and neurocognitive function between shift nurses and non-shift nurses. Methods: A cross-sectional descriptive study design was used. A total of 100 nurses participated in the study. 50 were shift nurses and the remaining 50 were non-shift nurses. The Pittsburgh Sleep Quality Index, Fatigue Severity Scale, and a computerized neurocognitive function test (CNS Vital Signs-VS4) were administered to the subjects to assess verbal and visual memory, processing speed, reaction time, and simple attention. After the last night shift, the shift nurse conducted the study at around 8:00 am and the non-shift nurse participated after work. Results: Compared to non-shift nurses, shift nurses had a significantly lower sleep quality (p=.002) and higher fatigue (p=.001) and achieved significantly lower scores on verbal memory (p=.001), processing speed (p=.003), and reaction time (p=.018). There were significant correlations between sleep quality and processing speed (p=.042), and reaction time (p=.015) of shift nurses who were bad sleepers. Conclusion: This study findings suggest shift work could interfere with cognitive function. Personal and organizational programs should be developed to support their sleep and neurocognitive function.

• Journal of Drive and Control
• /
• v.18 no.2
• /
• pp.46-55
• /
• 2021

A forklift is a powered industrial vehicle used to lift and move materials over short distances. Nowadays, almost all forklifts are equipped with an automatic transmission due to its improved operator comfort and increased productivity. Thanks to marked improvement of transmission control unit equipped with highly-advanced microcontrollers, recently developed automatic transmission for forklift have various auxiliary functions such as creep, auto retardation, and automatic shift with excellent shift quality. This paper deals with the creep function which enables one to maneuver a forklift at the designated low speed by slip control of clutches. The design of creep function was based on four modes of creep operation depending on the status of the operator's shift lever and accelerator pedal. Control algorithms and control parameters for each mode were designed to achieve the desired static and dynamic performance. Vehicle test for the designed creep function was carried out with an independently developed embedded controller. Test results confirmed good creep speed control without speed error at a steady state with a mild shift shock during mode changes by stepping or releasing the accelerator.

• Journal of the Korea Management Engineers Society
• /
• v.23 no.4
• /
• pp.87-103
• /
• 2018

This paper introduces the method to develop two main types of the fuzzy OEE (Overall Equipment Effectiveness) models via triangular membership function for measuring uncertainty. The fuzzy OEE includes model type 1 and model type 2. The model type 1 is used when the theoretical machine speed only reflects the time loss whereas model type 2 is used when the actual machine speed reflects both time and speed loss. Model type 2 has shown to perform a lower availability rate and a higher performance rate compared to model type 1. In addition, the fuzzy UPH (Unit Per Hour) which is derived from using the fuzzy OEE is presented to satisfy demand uncertainty. The fuzzy UPH can easily measure the fuzzy tact time and cycle time by reciprocating itself. Finally, this study demonstrates the fuzzy OEE models using IVIFS (Interval-Valued Intuitionistic Fuzzy Set) based on the characterization via membership function, non-membership function and hesitant function. For the purpose of analyzing the fuzzy system OEE, the OEE for each machine of plant structure is considered triangular interval-valued intuitionistic fuzzy number. Regardless of plant structure, the validity degree of fuzzy membership function of system OEE decreases when the number of machine with worst value of the validity degree increases. Corresponding examples are presented in this paper for practitioner to understand the applicability and practicability of the proposed fuzzy OEE methods.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.63 no.1
• /
• pp.11-25
• /
• 2021

The objective of this study was to analyze the impact of activation functions on flood forecasting model based on Artificial neural networks (ANNs). The traditional activation functions, the sigmoid and tanh functions, were compared with the functions which have been recently recommended for deep neural networks; the ReLU, leaky ReLU, and ELU functions. The flood forecasting model based on ANNs was designed to predict real-time runoff for 1 to 6-h lead time using the rainfall and runoff data of the past nine hours. The statistical measures such as R2, Nash-Sutcliffe Efficiency (NSE), Root Mean Squared Error (RMSE), the error of peak time (ETp), and the error of peak discharge (EQp) were used to evaluate the model accuracy. The tanh and ELU functions were most accurate with R2=0.97 and RMSE=30.1 (㎥/s) for 1-h lead time and R2=0.56 and RMSE=124.6~124.8 (㎥/s) for 6-h lead time. We also evaluated the learning speed by using the number of epochs that minimizes errors. The sigmoid function had the slowest learning speed due to the 'vanishing gradient problem' and the limited direction of weight update. The learning speed of the ELU function was 1.2 times faster than the tanh function. As a result, the ELU function most effectively improved the accuracy and speed of the ANNs model, so it was determined to be the best activation function for ANNs-based flood forecasting.

• The Mathematical Education
• /
• v.62 no.4
• /
• pp.473-490
• /
• 2023

Previous studies from domestic and abroad are accumulating information on how to reason students' continuous changes through teaching experiments. These studies deal with scenes in which students who make 'smooth reasoning' and 'chunky reasoning' construct mathematical results together in teaching experiments. However, in order to analyze their results in more detail, it is necessary to check what kind of results a student reasoning in a specific way constructs for the tasks of previous studies. According to the need for these studies, the researcher conducted a total of 14 teaching experiments on one first-year high school student who was found to make 'chunky reasoning'. In this study, it was possible to observe a scene in which a student who makes 'chunky reasoning' constructs an output similar to 'a mathematical result constructed by students with various reasoning methods(smooth reasnoning or chunky reasoning) in previous studies.' In particular, the student who participated in this study observed a consistent construction method of constructing the function of 'time-speed' from the function of 'time-distance'. The researcher expected that information on this student's distinctive construction methods would be helpful for subsequent studies.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.06a
• /
• pp.163-168
• /
• 2005

It is one of the most important things to determinate optimized cutting conditions which satisfy productivity and cost simultaneously in production and CAPP systems. These days many researchers have figured out the optimizing way for solutions of multi-object function to find the approach methods using algorithm such as genetic algorithm or tabu search, etc., instead of mathematical methods. The main creation of objective function is proposed by empirical method but which is difficult to set it up and to analysis. In this paper, an optimization method of cutting condition is shown using the ANN and GA for the multi-objective function in high speed machining.

• Proceedings of the Korea Committee for Ocean Resources and Engineering Conference
• /
• 2000.04a
• /
• pp.23-28
• /
• 2000

The Green integral equation for the calculation of the forward-speed time-harmonic radiation-diffraction potentials IS derived. The forward-speed Green function presented by Brard is used and the correct free surface boundary condition for the Green function is imposed. The cause of the mistakes in the existing Green integral equation is also pointed out.

• Proceedings of the KSR Conference
• /
• 1999.11a
• /
• pp.130-138
• /
• 1999

This study was carried out about interior design of trailer car for Korean High speed Train of Maximum operating speed of 350km/h. The interior main item design of high speed train is performed in order to satisfy the function of speed, safety and comfort concerning the RAMSH, ENVIRONMENT, ACTIVITY AND INFORMATION, This paper focuses on what is the design principle concerning the above mentioned factors and how did we apply that in the korea high speed train.