• Proceedings of the KSME Conference
• /
• 2001.11a
• /
• pp.533-538
• /
• 2001

This paper describes a combinational method to compute the global and local solutions of optimization problems. The present hybrid algorithm uses both a genetic algorithm and a local concentrate search algorithm (e. g simplex method). The hybrid algorithm is not only faster than the standard genetic algorithm but also supplies a more accurate solution. In addition, this algorithm can find the global and local optimum solutions. The present algorithm can be supplied to minimize the resonance response (Q factor) and to yield the critical speeds as far from the operating speed as possible. These factors play very important roles in designing a rotor-bearing system under the dynamic behavior constraint. In the present work, the shaft diameter, the bearing length, and clearance are used as the design variables.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.33 no.8
• /
• pp.753-759
• /
• 2009

Feed-forward neural networks have been widely used as function approximation tools in the context of global approximate optimization. In the present study, a wavelet neural network (WNN) which is based on wavelet transform theory is suggested as an alternative to a traditional back-propagation neural network (BPN). The basic theory of wavelet neural network is briefly described, and approximation performance is tested using a nonlinear multimodal function and a composite rotor blade analysis problem. Laplacian of Gaussian function, Mexican function, and Morlet function are considered during the construction of WNN architectures. In addition, approximation results from WNN are compared with those from BPN.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• International Journal of Computer Science & Network Security
• /
• v.22 no.10
• /
• pp.171-176
• /
• 2022

A dramatic rise in the number of people dying from heart disease has prompted efforts to find a way to identify it sooner using efficient approaches. A variety of variables contribute to the condition and even hereditary factors. The current estimate approaches use an automated diagnostic system that fails to attain a high level of accuracy because it includes irrelevant dataset information. This paper presents an effective neural network with convolutional layers for classifying clinical data that is highly class-imbalanced. Traditional approaches rely on massive amounts of data rather than precise predictions. Data must be picked carefully in order to achieve an earlier prediction process. It's a setback for analysis if the data obtained is just partially complete. However, feature extraction is a major challenge in classification and prediction since increased data increases the training time of traditional machine learning classifiers. The work integrates the CNN-MDRP classifier (convolutional neural network (CNN)-based efficient multimodal disease risk prediction with TANFIS (tuned adaptive neuro-fuzzy inference system) for earlier accurate prediction. Perform data cleaning by transforming partial data to informative data from the dataset in this project. The recommended TANFIS tuning parameters are then improved using a Laplace Gaussian mutation-based grasshopper and moth flame optimization approach (LGM²G). The proposed approach yields a prediction accuracy of 98.40 percent when compared to current algorithms.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.30 no.1
• /
• pp.87-94
• /
• 2017

Reliability analysis(RA) and Reliability-based design optimization(RBDO) require statistical modeling of input random variables, which is parametrically or nonparametrically determined based on experimental data. For the parametric method, goodness-of-fit (GOF) test and model selection method are widely used, and a sequential statistical modeling method combining the merits of the two methods has been recently proposed. Kernel density estimation(KDE) is often used as a nonparametric method, and it well describes a distribution function when the number of data is small or a density function has multimodal distribution. Although accurate statistical models are needed to obtain accurate RA and RBDO results, accurate statistical modeling is difficult when the number of data is small. In this study, the accuracy of two statistical modeling methods, SSM and KDE, were compared according to the number of data. Through numerical examples, the RA results using the input models modeled by two methods were compared, and appropriate modeling method was proposed according to the number of data.