• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.401-403
• /
• 2019

In this paper, we propose a generative adversarial networks (GAN) based text-to-image generating method. In many natural language processing tasks, which word expressions are determined by their term frequency -inverse document frequency scores. Word2Vec is a type of neural network model that, in the case of an unlabeled corpus, produces a vector that expresses semantics for words in the corpus and an image is generated by GAN training according to the obtained vector. Thanks to the understanding of the word we can generate higher and more realistic images. Our GAN structure is based on deep convolution neural networks and pixel recurrent neural networks. Comparing the generated image with the real image, we get about 88% similarity on the Oxford-102 flowers dataset.

• Archives of Craniofacial Surgery
• /
• v.22 no.5
• /
• pp.223-231
• /
• 2021

The field of artificial intelligence (AI) is rapidly advancing, and AI models are increasingly applied in the medical field, especially in medical imaging, pathology, natural language processing, and biosignal analysis. On the basis of these advances, telemedicine, which allows people to receive medical services outside of hospitals or clinics, is also developing in many countries. The mechanisms of deep learning used in medical AI include convolutional neural networks, residual neural networks, and generative adversarial networks. Herein, we investigate the possibility of using these AI methods in the field of craniofacial surgery, with potential applications including craniofacial trauma, congenital anomalies, and cosmetic surgery.

• Advances in aircraft and spacecraft science
• /
• v.9 no.5
• /
• pp.367-375
• /
• 2022

Modeling the properties of complex alloys such as nickel superalloys is an extremely challenging scientific and engineering task. The model should take into account a large number of uncorrelated factors, for many of which information may be missing or vague. The individual contribution of one or another chemical element out of a dozen possible ligants cannot be determined by traditional methods. Moreover, there are no general analytical models describing the influence of elements on the characteristics of alloys. Artificial neural networks are one of the few statistical modeling tools that can account for many implicit correlations and establish correspondences that cannot be identified by other more familiar mathematical methods. However, such networks require careful tuning to achieve high performance, which is time-consuming. Data preprocessing can make model training much easier and faster. This article focuses on combining physics-based deep network configuration and input data engineering to simulate the solvus temperature of nickel superalloys. The used deep artificial neural network shows good simulation results. Thus, this method of numerical simulation can be easily applied to such problems.

• Smart Structures and Systems
• /
• v.22 no.2
• /
• pp.175-183
• /
• 2018

Time-series data often contain one of the most valuable pieces of information in many fields including manufacturing. Because time-series data are relatively cheap to acquire, they (e.g., vibration signals) have become a crucial part of big data even in manufacturing shop floors. Recently, deep-learning models have shown state-of-art performance for analyzing big data because of their sophisticated structures and considerable computational power. Traditional models for a machinery-monitoring system have highly relied on features selected by human experts. In addition, the representational power of such models fails as the data distribution becomes complicated. On the other hand, deep-learning models automatically select highly abstracted features during the optimization process, and their representational power is better than that of traditional neural network models. However, the applicability of deep-learning models to the field of prognostics and health management (PHM) has not been well investigated yet. This study integrates the "residual fitting" mechanism inherently embedded in the wavelet transform into the convolutional neural network deep-learning structure. As a result, the architecture combines a signal smoother and classification procedures into a single model. Validation results from rotor vibration data demonstrate that our model outperforms all other off-the-shelf feature-based models.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.2
• /
• pp.504-519
• /
• 2023

Heart disease is becoming the top reason of death all around the world. Diagnosing cardiac illness is a difficult endeavor that necessitates both expertise and extensive knowledge. Machine learning (ML) is becoming gradually more important in the medical field. Most of the works have concentrated on the prediction of cardiac disease, however the precision of the results is minimal, and data integrity is uncertain. To solve these difficulties, this research creates an Integrated Accurate-Secure Heart Disease Prediction (IAS) Model based on Deep Convolutional Neural Networks. Heart-related medical data is collected and pre-processed. Secondly, feature extraction is processed with two factors, from signals and acquired data, which are further trained for classification. The Deep Convolutional Neural Networks (DCNN) is used to categorize received sensor data as normal or abnormal. Furthermore, the results are safeguarded by implementing an integrity validation mechanism based on the hash algorithm. The system's performance is evaluated by comparing the proposed to existing models. The results explain that the proposed model-based cardiac disease diagnosis model surpasses previous techniques. The proposed method demonstrates that it attains accuracy of 98.5 % for the maximum amount of records, which is higher than available classifiers.

• KIISE Transactions on Computing Practices
• /
• v.23 no.12
• /
• pp.698-702
• /
• 2017

Since the introduction of Deep Neural Networks to the Natural Language Processing field, two major approaches have been considered for modeling text. One method involved learning embeddings, i.e. the distributed representations containing abstract semantics of words or sentences, with the textual context. The other strategy consisted of composing the embeddings trained by the above to get embeddings of longer texts. However, most studies of the composition methods just adopt word embeddings without consideration of the optimal embedding unit and the optimal method of composition. In this paper, we conducted experiments to analyze the optimal embedding unit and the optimal composition method for modeling longer texts, such as documents. In addition, we suggest a new discourse-based composition to overcome the limitation of the sequential composition method on composing sentence embeddings.

• Journal of Convergence for Information Technology
• /
• v.10 no.8
• /
• pp.23-34
• /
• 2020

Recently, the artificial intelligence deep learning field has been hard to commercialize due to the high computing power and the price problem of computing resources. In this paper, we apply a double pruning techniques to evaluate the performance of the in-depth neural network and various datasets. Double pruning combines basic Network-slimming and Parameter-prunning. Our proposed technique has the advantage of reducing the parameters that are not important to the existing learning and improving the speed without compromising the learning accuracy. After training various datasets, the pruning ratio was increased to reduce the size of the model.We confirmed that MobileNet-V3 showed the highest performance as a result of NetScore performance analysis. We confirmed that the performance after pruning was the highest in MobileNet-V3 consisting of depthwise seperable convolution neural networks in the Cifar 10 dataset, and VGGNet and ResNet in traditional convolutional neural networks also increased significantly.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.8
• /
• pp.343-348
• /
• 2019

In recent years, extractive summarization systems based on end-to-end deep learning models have become popular. These systems do not require human-crafted features and adopt data-driven approaches. However, previous related studies have shown that linguistic analysis features such as part-of-speeches, named entities and word's frequencies are useful for extracting important sentences from a document to generate a summary. In this paper, we propose an extractive summarization system based on deep neural networks using conventional linguistic analysis features. In order to prove the usefulness of the linguistic analysis features, we compare the models with and without those features. The experimental results show that the model with the linguistic analysis features improves the Rouge-2 F1 score by 0.5 points compared to the model without those features.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.5
• /
• pp.1310-1338
• /
• 2023

As Internet of Things (IoT) applications and devices rapidly grow, cyber-attacks on IoT networks/systems also have an increasing trend, thus increasing the threat to security and privacy. Botnet is one of the threats that dominate the attacks as it can easily compromise devices attached to an IoT networks/systems. The compromised devices will behave like the normal ones, thus it is difficult to recognize them. Several intelligent approaches have been introduced to improve the detection accuracy of this type of cyber-attack, including deep learning and machine learning techniques. Moreover, dimensionality reduction methods are implemented during the preprocessing stage. This research work proposes deep Autoencoder dimensionality reduction method combined with Artificial Neural Network (ANN) classifier as botnet detection system for IoT networks/systems. Experiments were carried out using 3- layer, 4-layer and 5-layer pre-processing data from the MedBIoT dataset. Experimental results show that using a 5-layer Autoencoder has better results, with details of accuracy value of 99.72%, Precision of 99.82%, Sensitivity of 99.82%, Specificity of 99.31%, and F1-score value of 99.82%. On the other hand, the 5-layer Autoencoder model succeeded in reducing the dataset size from 152 MB to 12.6 MB (equivalent to a reduction of 91.2%). Besides that, experiments on the N_BaIoT dataset also have a very high level of accuracy, up to 99.99%.

• KIISE Transactions on Computing Practices
• /
• v.23 no.5
• /
• pp.299-303
• /
• 2017

By piling up hidden layers in artificial neural networks, deep learning is delivering outstanding performances for high-level abstraction problems such as object/speech recognition and natural language processing. Alternatively, deep-learning users often struggle with the tremendous amounts of time and resources that are required to train deep neural networks. To alleviate this computational challenge, many approaches have been proposed in a diversity of areas. In this work, two of the existing Apache Spark-based acceleration frameworks for deep learning (SparkNet and DeepSpark) are compared and analyzed in terms of the training accuracy and the time demands. In the authors' experiments with the CIFAR-10 and CIFAR-100 benchmark datasets, SparkNet showed a more stable convergence behavior than DeepSpark; but in terms of the training accuracy, DeepSpark delivered a higher classification accuracy of approximately 15%. For some of the cases, DeepSpark also outperformed the sequential implementation running on a single machine in terms of both the accuracy and the running time.