• Journal of Korean Society of Transportation
• /
• v.20 no.1
• /
• pp.131-139
• /
• 2002

This study introduces a developed artificial neural networks(ANN) model as a more efficient and reliable prediction model in operating speed Prediction with the 85th percentile horizontal curve of two-way rural highway in the aspect of evaluating highway design consistency. On the assumption that the speed is decided by highway geometry features, total 30 survey sites were selected. Data include currie radius, curve length, intersection angle, sight distance, lane width, and lane of those sites and were used as input layer data of the ANN. The optimized model structure was drawn by number of unit of hidden layer, learning coefficient, momentum coefficient, and change in learning frequency in multi-layer a ANN model. To verify learning Performance of ANN, 30 survey sites were selected while data in obtained from the 20 cites were used as learning data and those from the remaining 10 sites were used as predictive data. As a result of statistical verification, the model D of 4 types of ANN was evaluated as the most similar model to the actual operating speed value: R2 was 85% and %RMSE was 0.0204.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.5
• /
• pp.602-608
• /
• 2011

In this study, a method of designing a neurointerface using neural network (NN) is proposed for controlling nonholonomic mobile robots. According to the concept of virtual master-slave robots, in particular, a partially stable inverse dynamic model of the master robot is acquired online through the NN by applying a feedback-error learning method, in which the feedback controller is assumed to be based on a PD compensator for such a nonholonomic robot. The NN for the online feedback-error learning can composed that the input layer consists of six units for the inputs $x_i$, i=1~6, the hidden layer consists of two hidden units for hidden outputs $o_j$, j=1~2, and the output layer consists of two units for the outputs ${\tau}_k$, k=1~2. A tracking control problem is demonstrated by some simulations for a nonholonomic mobile robot with two-independent driving wheels. The initial q value was set to [0, 5, ${\pi}$].

• Journal of the Semiconductor & Display Technology
• /
• v.18 no.2
• /
• pp.48-52
• /
• 2019

Virtual metrology, which is one of APC techniques, is a method to predict characteristics of manufactured films using machine learning with saving time and resources. As the photoresist is no longer a mask material for use in high aspect ratios as the CD is reduced, hard mask is introduced to solve such problems. Among many types of hard mask materials, amorphous carbon layer(ACL) is widely investigated due to its advantages of high etch selectivity than conventional photoresist, high optical transmittance, easy deposition process, and removability by oxygen plasma. In this study, VM using different machine learning algorithms is applied to predict the thickness of ACL and trained models are evaluated which model shows best prediction performance. ACL specimens are deposited by plasma enhanced chemical vapor deposition(PECVD) with four different process parameters(Pressure, RF power, $C_3H_6$ gas flow, $N_2$ gas flow). Gradient boosting regression(GBR) algorithm, random forest regression(RFR) algorithm, and neural network(NN) are selected for modeling. The model using gradient boosting algorithm shows most proper performance with higher R-squared value. A model for predicting the thickness of the ACL film within the abovementioned conditions has been successfully constructed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.16 no.10
• /
• pp.934-940
• /
• 1991

A modified quantizing method is introduced to teach single layer learning algorithm, which is implemented optically. The proposed optical system consists of input masks, holographic diffraction grating. LCD and CCD camera. The 2 dimensional interconnections between input neurons and output neurons are realized using holographic phase grating, which is fabricated for equal intensity distribution of diffraction orders. The two gray levels of LCD act as binary weights for each interconnection. The weights are compensated according to the learning algorithm in which the amount of weights to be compensated is determined by comparing the output patterns with target patterns. The learning process is iterated until the predetermined conditions are satisfied. Optical experiments are performed for two learning rates, 0.5 and 0.9 and the experimental results show that the proposed system is useful for optical neural networks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.16 no.1
• /
• pp.35-45
• /
• 1991

Multi layer neural network using backpropagation learning algorithm is used to achieve identification and orientation estimation of different classes of aircraft in the variety of 3-D orientations. In-plane distortion invarient$(L,\;{\Phi})$ feature was extracted from each aircraft image to be used for training neural network aircraft classifier. For aircraft identification the optimum structure of the neural network classifier is studied to obtain high classification performance. Effective reductioin of learning time was achieved by using modified backpropagation learning algorithm and varying, learning parameters.

• International Journal of Contents
• /
• v.7 no.1
• /
• pp.14-22
• /
• 2011

Temporal Sign Language Recognition (TSLR) from hand motion is an active area of gesture recognition research in facilitating efficient communication with deaf people. TSLR systems consist of two stages: a motion sensing step which extracts useful features from signers' motion and a classification process which classifies these features as a performed sign. This work focuses on two of the research problems, namely unknown time varying signal of sign languages in feature extraction stage and computing complexity and time consumption in classification stage due to a very large sign sequences database. In this paper, we propose a combination of Dynamic Time Warping (DTW) and application of the Single hidden Layer Feedforward Neural networks (SLFNs) trained by Extreme Learning Machine (ELM) to cope the limitations. DTW has several advantages over other approaches in that it can align the length of the time series data to a same prior size, while ELM is a useful technique for classifying these warped features. Our experiment demonstrates the efficiency of the proposed method with the recognition accuracy up to 98.67%. The proposed approach can be generalized to more detailed measurements so as to recognize hand gestures, body motion and facial expression.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.4
• /
• pp.203-210
• /
• 2017

In this paper, we propose an effective neural network model for image caption generation and model transfer. This model is a kind of multi-modal recurrent neural network models. It consists of five distinct layers: a convolution neural network layer for extracting visual information from images, an embedding layer for converting each word into a low dimensional feature, a recurrent neural network layer for learning caption sentence structure, and a multi-modal layer for combining visual and language information. In this model, the recurrent neural network layer is constructed by LSTM units, which are well known to be effective for learning and transferring sequence patterns. Moreover, this model has a unique structure in which the output of the convolution neural network layer is linked not only to the input of the initial state of the recurrent neural network layer but also to the input of the multimodal layer, in order to make use of visual information extracted from the image at each recurrent step for generating the corresponding textual caption. Through various comparative experiments using open data sets such as Flickr8k, Flickr30k, and MSCOCO, we demonstrated the proposed multimodal recurrent neural network model has high performance in terms of caption accuracy and model transfer effect.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.4
• /
• pp.60-68
• /
• 1994

In this paper we propose a new Hybrid learning rule applied to multilayer feedforward neural networks, which is constructed by combining Hebbian learning rule that is a good feature extractor and Back-Propagation(BP) learning rule that is an excellent classifier. Unlike the BP rule used in multi-layer perceptron(MLP), the proposed Hybrid learning rule is used for uptate of all connection weights except for output connection weigths becase the Hebbian learning in output layer does not guarantee learning convergence. To evaluate the performance, the proposed hybrid rule is applied to classifier problems in two dimensional space and shows better performance than the one applied only by the BP rule. In terms of learning speed the proposed rule converges faster than the conventional BP. For example, the learning of the proposed Hybrid can be done in 2/10 of the iterations that are required for BP, while the recognition rate of the proposed Hybrid is improved by about $0.778\%$ at the peak.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.1017-1020
• /
• 1993

This paper describes a neo fuzzy neuron which was produced by a fusion of fuzzy logic and neuroscience. Some learning algorithms are presented. The guarantee for the global minimum on the error-weight space is proved by a reduction to absurdity. Enhanced is that the learning speed of the neo fuzzy neuron exceeds 100,000 times of that of conventional multi-layer neural networks.

• Journal of Digital Policy
• /
• v.2 no.4
• /
• pp.25-32
• /
• 2023

This paper is a study to develop a deep neural network (DNN) blood glucose prediction model based on heart rate (HR) and heart rate variability (HRV) data measured by PPG-based sensors. MLP deep learning consists of an input layer, a hidden layer, and an output layer with 11 independent variables. The learning results of the blood glucose prediction model are MAE=0.3781, MSE=0.8518, and RMSE=0.9229, and the coefficient of determination (R²) is 0.9994. The study was able to verify the feasibility of glycemic control using non-blood vital signs using PPG-based digital devices. In conclusion, a standardized method of acquiring and interpreting PPG-based vital signs, a large data set for deep learning, and a study to demonstrate the accuracy of the method may provide convenience and an alternative method for blood glucose management in dogs.