• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.26 no.4
• /
• pp.402-409
• /
• 2017

Selection of optimal cutting conditions is important for improving productivity and implementing efficient process control in metal machining. In this study, improvement of cutting conditions in machining using end-mills is studied by using deep-layered neural networks, which comprise an input layer, output layer, and two hidden layers. System networks are designed with inputs as cutting conditions, and they output the cutting force. A pseudo-inverse network is designed that has the adjustable cutting condition as output and cutting force and other cutting conditions as input. The combination of the system network and pseudo-inverse network enables selection or improvement of cutting conditions that results in the expected cutting force.

• Structural Engineering and Mechanics
• /
• v.77 no.1
• /
• pp.47-56
• /
• 2021

This paper deals with damage detection using a Gapped Smoothing Method (GSM) combined with deep learning. Convolutional Neural Network (CNN) is a model of deep learning. CNN has an input layer, an output layer, and a number of hidden layers that consist of convolutional layers. The input layer is a tensor with shape (number of images) × (image width) × (image height) × (image depth). An activation function is applied each time to this tensor passing through a hidden layer and the last layer is the fully connected layer. After the fully connected layer, the output layer, which is the final layer, is predicted by CNN. In this paper, a complete machine learning system is introduced. The training data was taken from a Finite Element (FE) model. The input images are the contour plots of curvature gapped smooth damage index. A free-free beam is used as a case study. In the first step, the FE model of the beam was used to generate data. The collected data were then divided into two parts, i.e. 70% for training and 30% for validation. In the second step, the proposed CNN was trained using training data and then validated using available data. Furthermore, a vibration experiment on steel damaged beam in free-free support condition was carried out in the laboratory to test the method. A total number of 15 accelerometers were set up to measure the mode shapes and calculate the curvature gapped smooth of the damaged beam. Two scenarios were introduced with different severities of the damage. The results showed that the trained CNN was successful in detecting the location as well as the severity of the damage in the experimental damaged beam.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.3 no.2
• /
• pp.313-319
• /
• 1999

We devise the 3 models of Multilayered Recurrent Prediction Neural Network(MLRPNN), which are obtained by modifying the Multilayered Perceptron(MLP) with 4 layers. We experimentally study the speech recognition performance of 3 models by a comparative method, according to the variation of the prediction order, the number of neurons in two hidden layers, initial values of connecting weights and transfer function, respectively. By the experiment, the recognition performance of each MLRPNN is better than that of MLP. At the model that returns the output of the upper hidden layer to the lower hidden layer, the recognition performance shows the best value. All MLRPNNs, which have 10 or 15 neurons in the upper and lower hidden layer and is predicted by 3rd or 4th order, show the improved speech recognition rate. On learning, these MLRPNNs have a better recognition rate when we set the initial weights between -0.5 and 0.5, and use the unipolar sigmoid transfer function in the lower hidden layer.

• Journal of Practical Engineering Education
• /
• v.15 no.2
• /
• pp.381-388
• /
• 2023

Through artificial neural network education using spreadsheets, non-major undergraduate students can understand the operation principle of artificial neural networks and develop their own artificial neural network software. Here, training of the operation principle of artificial neural networks starts with the generation of training data and the assignment of correct answer labels. Then, the output value calculated from the firing and activation function of the artificial neuron, the parameters of the input layer, hidden layer, and output layer is learned. Finally, learning the process of calculating the error between the correct label of each initially defined training data and the output value calculated by the artificial neural network, and learning the process of calculating the parameters of the input layer, hidden layer, and output layer that minimize the total sum of squared errors. Training on the operation principles of artificial neural networks using a spreadsheet was conducted for undergraduate non-major students. And image training data and basic artificial neural network development results were collected. In this paper, we analyzed the results of collecting two types of training data and the corresponding artificial neural network SW with small 12-pixel images, and presented methods and execution results of using the collected training data for Orange machine learning model learning and analysis tools.

• Asian Journal of Atmospheric Environment
• /
• v.10 no.2
• /
• pp.67-79
• /
• 2016

The forecasting of air pollution is an important and popular topic in environmental engineering. Due to health impacts caused by unacceptable particulate matter (PM) levels, it has become one of the greatest concerns in metropolitan cities like Karaj City in Iran. In this study, the concentration of $PM_{2.5}$ was predicted by applying a multilayer percepteron (MLP) neural network, a radial basis function (RBF) neural network and a Markov chain model. Two months of hourly data including temperature, NO, $NO_2$, $NO_x$, CO, $SO_2$ and $PM_{10}$ were used as inputs to the artificial neural networks. From 1,488 data, 1,300 of data was used to train the models and the rest of the data were applied to test the models. The results of using artificial neural networks indicated that the models performed well in predicting $PM_{2.5}$ concentrations. The application of a Markov chain described the probable occurrences of unhealthy hours. The MLP neural network with two hidden layers including 19 neurons in the first layer and 16 neurons in the second layer provided the best results. The coefficient of determination ($R^2$), Index of Agreement (IA) and Efficiency (E) between the observed and the predicted data using an MLP neural network were 0.92, 0.93 and 0.981, respectively. In the MLP neural network, the MBE was 0.0546 which indicates the adequacy of the model. In the RBF neural network, increasing the number of neurons to 1,488 caused the RMSE to decline from 7.88 to 0.00 and caused $R^2$ to reach 0.93. In the Markov chain model the absolute error was 0.014 which indicated an acceptable accuracy and precision. We concluded the probability of occurrence state duration and transition of $PM_{2.5}$ pollution is predictable using a Markov chain method.

• Advances in environmental research
• /
• v.5 no.3
• /
• pp.153-167
• /
• 2016

We applied multilayer perceptron (MLP) and radial basis function (RBF) neural network in upstream and downstream water quality stations of the Karaj Reservoir in Iran. For both neural networks, inputs were pH, turbidity, temperature, chlorophyll-a, biochemical oxygen demand (BOD) and nitrate, and the output was dissolved oxygen (DO). We used an MLP neural network with two hidden layers, for upstream station 15 and 33 neurons in the first and second layers respectively, and for the downstream station, 16 and 21 neurons in the first and second hidden layer were used which had minimum amount of errors. For learning process 6-fold cross validation were applied to avoid over fitting. The best results acquired from RBF model, in which the mean bias error (MBE) and root mean squared error (RMSE) were 0.063 and 0.10 for the upstream station. The MBE and RSME were 0.0126 and 0.099 for the downstream station. The coefficient of determination ($R^2$) between the observed data and the predicted data for upstream and downstream stations in the MLP was 0.801 and 0.904, respectively, and in the RBF network were 0.962 and 0.97, respectively. The MLP neural network had acceptable results; however, the results of RBF network were more accurate. A sensitivity analysis for the MLP neural network indicated that temperature was the first parameter, pH the second and nitrate was the last factor affecting the prediction of DO concentrations. The results proved the workability and accuracy of the RBF model in the prediction of the DO.

• Journal of the Korean Graphic Arts Communication Society
• /
• v.15 no.1
• /
• pp.31-40
• /
• 1997

The objectives of color reproduction in printing, photography, and digital hard-copy is an important problem. The Color is obsorved differently from illumination an obsorvation condition, and varied according to individual taste. Generally, the color reproduction system is designed with colorimetric color reproduction method. But the color gamut of the color reproduction system is different each other and the one device has nonlinear relationalship between the other. By these reason, to predict the reproduced color based on linear color transform method is difficult. Some methods of non-linear color transform by neural network was proposed. These method was theoretical useful and valid to transform from CIE color to device color. But more studies were needed to realize the non-linear color transform system. In this paper, we described a method to realize the non-linear color transform system by neural network. The optimum structure of the non-linear color transform system was found out. The structure of descrived system has four layer( input, output and two hidden layers.) Input and output layer have 3 units, and a hidden layer has 27 units. We trained 216 color-samples, and estimated the realized color transform system by 1115 color-samples. The average color difference between original color samples and transformed color samples was 2.54.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.170-170
• /
• 2020

Due to the development of technologies, complex computation of huge data set is possible with a prevalent personal computer. Therefore, machine learning methods have been widely applied in the hydrologic field such as regression-based regional frequency analysis (RFA). The main purpose of this study is to compare two frameworks of RFA based on the artificial neural network (ANN) models: quantile regression technique (QRT-ANN) and parameter regression technique (PRT-ANN). As an output layer of the ANN model, the QRT-ANN predicts quantiles for various return periods whereas the PRT-ANN provides prediction of three parameters for the generalized extreme value distribution. Rainfall gauging sites where record length is more than 20 years were selected and their annual maximum rainfalls and various hydro-meteorological variables were used as an input layer of the ANN model. While employing the ANN model, 70% and 30% of gauging sites were used as training set and testing set, respectively. For each technique, ANN model structure such as number of hidden layers and nodes was determined by a leave-one-out validation with calculating root mean square error (RMSE). To assess the performances of two frameworks, RMSEs of quantile predicted by the QRT-ANN are compared to those of the PRT-ANN.

• Journal of Intelligence and Information Systems
• /
• v.3 no.1
• /
• pp.47-68
• /
• 1997

This paper discusses a method of modified Elman network(1990) for nonlinear predictions and its a, pp.ication to forecasting daily exchange rate returns. The method consists of two stages that take advantages of both time domain filter and modified feedback networks. The first stage straightforwardly employs the filtering technique to remove extreme noise. In the second stage neural networks are designed to take the feedback from both hidden-layer units and the deviation of outputs from target values during learning. This combined feedback can be exploited to transfer unconsidered information on errors into the network system and, consequently, would improve predictions. The method a, pp.ars to dominate linear ARMA models and standard dynamic neural networks in one-step-ahead forecasting exchange rate returns.

• Earthquakes and Structures
• /
• v.15 no.4
• /
• pp.431-441
• /
• 2018

This study provides fragility-based assessment of seismic performance of reinforced concrete bridges. Seismic fragility curves were created using nonlinear analysis (NA) and artificial neural networks (ANNs). Nonlinear response history analyses were performed, in order to calculate the seismic performances of the bridges. To this end, 306 bridge-earthquake cases were considered. A multi-layered perceptron (MLP) neural network was implemented to predict the seismic performances of the selected bridges. The MLP neural networks considered herein consist of an input layer with four input vectors; two hidden layers and an output vector. In order to train ANNs, 70% of the numerical results were selected, and the remained 30% were employed for testing the reliability and validation of ANNs. Several structures of MLP neural networks were examined in order to obtain suitable neural networks. After achieving the most proper structure of neural network, it was used for generating new data. A total number of 600 new bridge-earthquake cases were generated based on neural simulation. Finally, probabilistic seismic safety analyses were conducted. Herein, fragility curves were developed using numerical results, neural predictions and the combination of numerical and neural data. Results of this study revealed that ANNs are suitable tools for predicting seismic performances of RC bridges. It was also shown that yield stresses of the reinforcements is one of the important sources of uncertainty in fragility analysis of RC bridges.