• Journal of Internet Computing and Services
• /
• v.21 no.3
• /
• pp.83-92
• /
• 2020

Recently, many companies and organizations are interested in predictive process monitoring for the efficient operation of business process models. Traditional process monitoring focused on the elapsed execution state of a particular process instance. On the other hand, predictive process monitoring focuses on predicting the future execution status of a particular process instance. In this paper, we implement the function of the business process remaining time prediction, which is one of the predictive process monitoring functions. In order to effectively model the remaining time, normalization by activity is proposed and applied to the predictive model by taking into account the difference in the distribution of time feature values according to the properties of each activity. In order to demonstrate the superiority of the predictive performance of the proposed model in this paper, it is compared with previous studies through event log data of actual companies provided by 4TU.Centre for Research Data.

• KIISE Transactions on Computing Practices
• /
• v.23 no.5
• /
• pp.316-321
• /
• 2017

We analyze the online Hangul handwriting recognition problem (HHR) and present solutions based on recurrent neural networks. The solutions are organized according to the three kinds of sequence labeling problem - sequence classifications, segment classification, and temporal classification, with additional consideration of the structural constitution of Hangul characters. We present a stacked gated recurrent unit (GRU) based model as the natural HHR solution in the sequence classification level. The proposed model shows 86.2% accuracy for recognizing 2350 Hangul characters and 98.2% accuracy for recognizing the six types of Hangul characters. We show that the type recognizing model successfully follows the type change as strokes are sequentially written. These results show the potential for RNN models to learn high-level structural information from sequential data.

• The Korean Journal of Applied Statistics
• /
• v.32 no.6
• /
• pp.899-908
• /
• 2019

Massive in-depth neural networks with numerous parameters are powerful machine learning methods, but they have overfitting problems due to the excessive flexibility of the models. Dropout is one methods to overcome the problem of oversized neural networks. It is also an effective method that randomly drops input and hidden nodes from the neural network during training. Every sample is fed to a thinned network from an exponential number of different networks. In this study, instead of feeding one sample for each thinned network, two or more samples are used in fitting for one thinned network known as a Hybrid Dropout. Simulation results using real data show that the new method improves the stability of estimates and reduces the minimum error for the verification data.

• Journal of Korea Society of Industrial Information Systems
• /
• v.22 no.6
• /
• pp.23-29
• /
• 2017

Due to the success of Wealthfront, Betterment, etc., there is a growing interest in RoboAdvisor that is an automated asset allocation methodology globally. RoboAdvisor minimizes human involvement in managing assets, thereby reducing the costs of using services and eliminating human psychological factors. In this paper, we developed a predictive model for the KOSPI 200 Futures Index using deep learning, in order to replace the existing technical analysis technique. And the proposed model confirmed that When the KOSPI 200 Gift Index is small, it can be used to predict direction and price of index. In combination with the existing technical analysis, It is confirmed that the proposed models combining with existing technical analyses and can be applied to the RoboAdvisor Service in the future.

• Korean Journal of Remote Sensing
• /
• v.36 no.4
• /
• pp.515-525
• /
• 2020

This study investigates the potential of bidirectional long short-term memory (Bi-LSTM) for efficient modeling of temporal information in crop classification using multitemporal remote sensing images. Unlike unidirectional LSTM models that consider only either forward or backward states, Bi-LSTM could account for temporal dependency of time-series images in both forward and backward directions. This property of Bi-LSTM can be effectively applied to crop classification when it is difficult to obtain full time-series images covering the entire growth cycle of crops. The classification performance of the Bi-LSTM is compared with that of two unidirectional LSTM architectures (forward and backward) with respect to different input image combinations via a case study of crop classification in Anbadegi, Korea. When full time-series images were used as inputs for classification, the Bi-LSTM outperformed the other unidirectional LSTM architectures; however, the difference in classification accuracy from unidirectional LSTM was not substantial. On the contrary, when using multitemporal images that did not include useful information for the discrimination of crops, the Bi-LSTM could compensate for the information deficiency by including temporal information from both forward and backward states, thereby achieving the best classification accuracy, compared with the unidirectional LSTM. These case study results indicate the efficiency of the Bi-LSTM for crop classification, particularly when limited input images are available.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.39-46
• /
• 2021

This study is about a method of extracting a summary from a news article in consideration of the importance of each sentence constituting the article. We propose a method of calculating sentence importance by extracting the probabilities of topic sentence, similarity with article title and other sentences, and sentence position as characteristics that affect sentence importance. At this time, a hypothesis is established that the Topic Sentence will have a characteristic distinct from the general sentence, and a deep learning-based classification model is trained to obtain a topic sentence probability value for the input sentence. Also, using the pre-learned ELMo language model, the similarity between sentences is calculated based on the sentence vector value reflecting the context information and extracted as sentence characteristics. The topic sentence classification performance of the LSTM and BERT models was 93% accurate, 96.22% recall, and 89.5% precision, resulting in high analysis results. As a result of calculating the importance of each sentence by combining the extracted sentence characteristics, it was confirmed that the performance of extracting the topic sentence was improved by about 10% compared to the existing TextRank algorithm.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.27 no.1
• /
• pp.34-40
• /
• 2023

A smart business card printing system that automatically prints business cards requested by customers online is being activated. What matters is that the business card submitted by the customer to the system may be abnormal. This paper deals with the problem of determining whether the image of a business card has been abnormally rotated by adopting artificial intelligence technology. It is assumed that the business card rotates 0 degrees, 90 degrees, 180 degrees, and 270 degrees. Experiments were conducted by applying existing VGG, ResNet, and DenseNet artificial neural networks without designing special artificial neural networks, and they were able to distinguish image rotation with an accuracy of about 97%. DenseNet161 showed 97.9% accuracy and ResNet34 also showed 97.2% precision. This illustrates that if the problem is simple, it can produce sufficiently good results even if the neural network is not a complex one.

• Journal of Software Assessment and Valuation
• /
• v.17 no.1
• /
• pp.75-82
• /
• 2021

Mastitis in cows is a major factor that hinders dairy productivity of farms, and many attempts have been made to solve it. However, research on mastitis has been limited to diagnosis rather than prediction, and even this is mostly using a single sensor. In this study, a predictive model was developed using multivariate data including biometric data and environmental data. The data used for the analysis were collected from robot milking machines and sensors installed in farmhouses in Chungcheongnam-do, South Korea. The recurrent neural network model using three weeks of data predicts whether or not mastitis is diagnosed the next day. As a result, mastitis was predicted with an accuracy of 82.9%. The superiority of the model was confirmed by comparing the performance of various data collection periods and various models.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.6
• /
• pp.1081-1090
• /
• 2022

The face recognition model is used for identity recognition of smartphones, providing convenience to many users. As a result, the security review of the DNN model is becoming important, with adversarial attacks present as a well-known vulnerability of the DNN model. Adversarial attacks have evolved to decision-based attack techniques that use only the recognition results of deep learning models to perform attacks. However, existing decision-based attack technique[14] have a problem that requires a large number of queries when generating adversarial examples. In particular, it takes a large number of queries to approximate the gradient. Therefore, in this paper, we propose a method of generating adversarial examples using orthogonal space sampling and dimensionality reduction sampling to avoid wasting queries that are consumed to approximate the gradient of existing decision-based attack technique[14]. Experiments show that our method can reduce the perturbation size of adversarial examples by about 2.4 compared to existing attack technique[14] and increase the attack success rate by 14% compared to existing attack technique[14]. Experimental results demonstrate that the adversarial example generation method proposed in this paper has superior attack performance.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.4
• /
• pp.577-586
• /
• 2022

Detection and classification of steel surface defects are critical for product quality control in the steel industry. However, due to its low accuracy and slow speed, the traditional approach cannot be effectively used in a production line. The current, widely used algorithm (based on deep learning) has an accuracy problem, and there are still rooms for development. This paper proposes a method of steel surface defect detection combining EfficientNetV2 for image classification and YOLOv5 as an object detector. Shorter training time and high accuracy are advantages of this model. Firstly, the image input into EfficientNetV2 model classifies defect classes and predicts probability of having defects. If the probability of having a defect is less than 0.25, the algorithm directly recognizes that the sample has no defects. Otherwise, the samples are further input into YOLOv5 to accomplish the defect detection process on the metal surface. Experiments show that proposed model has good performance on the NEU dataset with an accuracy of 98.3%. Simultaneously, the average training speed is shorter than other models.