• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.131-131
• /
• 2021

Deep learning models, especially those based on long short-term memory (LSTM), have presented their superiority in addressing time series data issues recently. This study aims to comprehensively evaluate the performance of deep learning models that belong to the supervised learning category in streamflow prediction. Therefore, six deep learning models-standard LSTM, standard gated recurrent unit (GRU), stacked LSTM, bidirectional LSTM (BiLSTM), feed-forward neural network (FFNN), and convolutional neural network (CNN) models-were of interest in this study. The Red River system, one of the largest river basins in Vietnam, was adopted as a case study. In addition, deep learning models were designed to forecast flowrate for one- and two-day ahead at Son Tay hydrological station on the Red River using a series of observed flowrate data at seven hydrological stations on three major river branches of the Red River system-Thao River, Da River, and Lo River-as the input data for training, validation, and testing. The comparison results have indicated that the four LSTM-based models exhibit significantly better performance and maintain stability than the FFNN and CNN models. Moreover, LSTM-based models may reach impressive predictions even in the presence of upstream reservoirs and dams. In the case of the stacked LSTM and BiLSTM models, the complexity of these models is not accompanied by performance improvement because their respective performance is not higher than the two standard models (LSTM and GRU). As a result, we realized that in the context of hydrological forecasting problems, simple architectural models such as LSTM and GRU (with one hidden layer) are sufficient to produce highly reliable forecasts while minimizing computation time because of the sequential data nature.

• Advances in nano research
• /
• v.15 no.4
• /
• pp.315-328
• /
• 2023

The main goal of the present study was to assess the effects of eggshell powder (ESP) and halloysite nanotubes (HNTs) on the mechanical properties of abaca fiber (AF)-reinforced natural composites. For this purpose, a limited number of indentation tests were first performed on the AF/polypropylene (PP) composites for different HNT and ESP loadings (0 wt.% ~ 6 wt.%), load amplitudes (150, 200, and 250 N), and two types of indenters (Vickers or conical). The Young's modulus, hardness and plasticity index of each specimen were calculated using the indentation test results and Oliver-Pharr method. The accuracy of the experimental results was confirmed by comparing the values of the Young's modulus obtained from the indentation test with the results of the conventional tensile test. Then, a feed-forward shallow artificial neural network (ANN) with high efficiency was trained based on the obtained experimental data. The trained ANN could properly predict the variations of the mentioned mechanical properties of AF/PP composites incorporated with different HNT and ESP loadings. Furthermore, the trained ANN demonstrated that HNTs increase the elastic modulus and hardness of the composite, while the incorporation of ESP reduces these properties. For instance, the Young's modulus of composites incorporated with 3 wt.% of ESP decreased by 30.7% compared with the pure composite, while increasing the weight fraction of ESP up to 6% decreased the Young's modulus by 34.8%. Moreover, the trained ANN indicated that HNTs have a more significant effect on reducing the plasticity index than ESP.

• IEIE Transactions on Smart Processing and Computing
• /
• v.2 no.6
• /
• pp.323-331
• /
• 2013

An extreme learning machine (ELM) is an efficient learning algorithm that is based on the generalized single, hidden-layer feed-forward networks (SLFNs), which perform well in classification applications. Many studies have demonstrated its superiority over the existing classical algorithms: support vector machine (SVM) and BP neural network. This paper presents a novel face recognition approach based on a multi-class project extreme learning machine (MPELM) classifier and 2D Gabor transform. First, all face image features were extracted using 2D Gabor filters, and the MPELM classifier was used to determine the final face classification. Two well-known face databases (CMU-PIE and ORL) were used to evaluate the performance. The experimental results showed that the MPELM-based method outperformed the ELM-based method as well as other methods.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.06a
• /
• pp.288-295
• /
• 1998

This paper proposes a new data classification method based on the tolerant rough set that extends the existing equivalent rough set. Similarity measure between two data is described by a distance function of all constituent attributes and they are defined to be tolerant when their similarity measure exceeds a similarity threshold value. The determination of optimal similarity theshold value is very important for the accurate classification. So, we determine it optimally by using the genetic algorithm (GA), where the goal of evolution is to balance two requirements such that (1) some tolerant objects are required to be included in the same class as many as possible. After finding the optimal similarity threshold value, a tolerant set of each object is obtained and the data set is grounded into the lower and upper approximation set depending on the coincidence of their classes. We propose a two-stage classification method that all data are classified by using the lower approxi ation at the first stage and then the non-classified data at the first stage are classified again by using the rough membership functions obtained from the upper approximation set. We apply the proposed classification method to the handwritten numeral character classification. problem and compare its classification performance and learning time with those of the feed forward neural network's back propagation algorithm.

• Environmental Engineering Research
• /
• v.24 no.3
• /
• pp.404-411
• /
• 2019

In this study, we designed a data-driven model to predict chlorophyll-a using M5P model tree and extreme learning machine (ELM). The Juksan weir in the Youngsan River has high chlorophyll-a, which is the primary indicator of algal bloom every year. Short-term algal bloom prediction is important for environmental management and ecological assessment. Two models were developed and evaluated for short-term algal bloom prediction. M5P is a classification and regression-analysis-based method, and ELM is a feed-forward neural network with fast learning using the least square estimate for regression. The dataset used in this study includes water temperature, rainfall, solar radiation, total nitrogen, total phosphorus, N/P ratio, and chlorophyll-a, which were collected on a daily basis from January 2013 to December 2016. The M5P model showed that the prediction model after one day had the highest performance power and dropped off rapidly starting with predictions after three days. Comparing the performance power of the ELM model with the M5P model, it was found that the performance power of the 1-7 d chlorophyll-a prediction model was higher. Moreover, in a period of rapidly increasing algal blooms, the ELM model showed higher accuracy than the M5P model.

• Journal of Veterinary Clinics
• /
• v.23 no.2
• /
• pp.105-110
• /
• 2006

A three-layer, feed-forward artificial neural network (ANN) with sixteen input neurons, three hidden neurons, and one output neuron was developed to identify the presence of infectious bronchitis (IB) infection as early as possible in laying hen flocks. Retrospective data from flocks that enrolled IB surveillance program between May 2003 and November 2005 were used to build the ANN. Data set of 86 flocks was divided randomly into two sets: 77 cases for training set and 9 cases for testing set. Input factors were 16 epidemiological findings including characteristics of the layer house, management practice, flock size, and the output was either presence or absence of IB. ANN was trained using training set with a back-propagation algorithm and test set was used to determine the network's capability to predict outcomes that it has never seen. Diagnostic performance of the trained network was evaluated by constructing receiver operating characteristic (ROC) curve with the area under the curve (AUC), which were also used to determine the best positivity criterion for the model. Several different ANNs with different structures were created. The best-fitted trained network, IBV_D1, was able to predict IB in 73 cases out of 77 (diagnostic accuracy 94.8%) in the training set. Sensitivity and specificity of the trained neural network was 95.5% (42/44, 95% CI, 84.5-99.4) and 93.9% (31/33, 95% CI, 79.8-99.3), respectively. For testing set, AVC of the ROC curve for the IBV_D1 network was 0.948 (SE=0.086, 95% CI 0.592-0.961) in recognizing IB infection status accurately. At a criterion of 0.7149, the diagnostic accuracy was the highest with a 88.9% with the highest sensitivity of 100%. With this value of sensitivity and specificity together with assumed 44% of IB prevalence, IBV_D1 network showed a PPV of 80% and an NPV of 100%. Based on these findings, the authors conclude that neural network can be successfully applied to the development of a screening model for identifying IB infection in laying hen flocks.

• International Journal of Fluid Machinery and Systems
• /
• v.8 no.4
• /
• pp.264-273
• /
• 2015

This presentation describes an experimental approach for the detection of cavitation in hydraulic machines by use of ultrasonic signal analysis. Instead of using the high frequency pulses (typically 1MHz) only for transit time measurement different other signal characteristics are extracted from the individual signals and its correlation function with reference signals in order to gain knowledge of the water conditions. As the pulse repetition rate is high (typically 100Hz), statistical parameters can be extracted of the signals. The idea is to find patterns in the parameters by a classifier that can distinguish between the different water states. This classification scheme has been applied to different cavitation sections: a sphere in a water flow in circular tube at the HSLU in Lucerne, a NACA profile in a cavitation tunnel and two Francis model test turbines all at LMH in Lausanne. From the signal raw data several statistical parameters in the time and frequency domain as well as from the correlation function with reference signals have been determined. As classifiers two methods were used: neural feed forward networks and decision trees. For both classification methods realizations with lowest complexity as possible are of special interest. It is shown that two to three signal characteristics, two from the signal itself and one from the correlation function are in many cases sufficient for the detection capability. The final goal is to combine these results with operating point, vibration, acoustic emission and dynamic pressure information such that a distinction between dangerous and not dangerous cavitation is possible.

• Smart Structures and Systems
• /
• v.22 no.5
• /
• pp.561-574
• /
• 2018

In this paper, for efficiently reducing the computational cost of the model updating during the optimization process of damage detection, the structural response is evaluated using properly trained surrogate model. Furthermore, in practice uncertainties in the FE model parameters and modelling errors are inevitable. Hence, an efficient approach based on Monte Carlo simulation is proposed to take into account the effect of uncertainties in developing a surrogate model. The probability of damage existence (PDE) is calculated based on the probability density function of the existence of undamaged and damaged states. The current work builds a framework for Probability Based Damage Detection (PBDD) of structures based on the best combination of metaheuristic optimization algorithm and surrogate models. To reach this goal, three popular metamodeling techniques including Cascade Feed Forward Neural Network (CFNN), Least Square Support Vector Machines (LS-SVMs) and Kriging are constructed, trained and tested in order to inspect features and faults of each algorithm. Furthermore, three wellknown optimization algorithms including Ideal Gas Molecular Movement (IGMM), Particle Swarm Optimization (PSO) and Bat Algorithm (BA) are utilized and the comparative results are presented accordingly. Furthermore, efficient schemes are implemented on these algorithms to improve their performance in handling problems with a large number of variables. By considering various indices for measuring the accuracy and computational time of PBDD process, the results indicate that combination of LS-SVM surrogate model by IGMM optimization algorithm have better performance in predicting the of damage compared with other methods.

• Nuclear Engineering and Technology
• /
• v.54 no.1
• /
• pp.36-48
• /
• 2022

As a valid numerical method to obtain a high-resolution result of a flow field, computational fluid dynamics (CFD) have been widely used to study coolant flow and heat transfer characteristics in fuel rod bundles. However, the time-consuming, iterative calculation of Navier-Stokes equations makes CFD unsuitable for the scenarios that require efficient simulation such as sensitivity analysis and uncertainty quantification. To solve this problem, a reduced-order model (ROM) based on proper orthogonal decomposition (POD) and machine learning (ML) is proposed to simulate the flow field efficiently. Firstly, a validated CFD model to output the flow field data set of the rod bundle is established. Secondly, based on the POD method, the modes and corresponding coefficients of the flow field were extracted. Then, an deep feed-forward neural network, due to its efficiency in approximating arbitrary functions and its ability to handle high-dimensional and strong nonlinear problems, is selected to build a model that maps the non-linear relationship between the mode coefficients and the boundary conditions. A trained surrogate model for modes coefficients prediction is obtained after a certain number of training iterations. Finally, the flow field is reconstructed by combining the product of the POD basis and coefficients. Based on the test dataset, an evaluation of the ROM is carried out. The evaluation results show that the proposed POD-ROM accurately describe the flow status of the fluid field in rod bundles with high resolution in only a few milliseconds.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.