• Journal of IKEEE
• /
• v.20 no.3
• /
• pp.209-219
• /
• 2016

SVM is a machine learning method used for image processing. It is well known for its high classification performance. We have to perform multiple MAC operations in order to use SVM for image classification. However, if the resolution of the target image or the number of classification cases increases, the execution time of SVM also increases, which makes it difficult to be performed in real-time applications. In this paper, we propose an hardware architecture which enables real-time applications using SVM classification. We used parallel architecture to simultaneously calculate MAC operations, and also designed the system for several feature extractors for compatibility. RBF kernel was used for hardware implemenation, and the exponent calculation formular included in the kernel was modified to enable fixed point modelling. Experimental results for the system, when implemented in Xilinx ZC-706 evaluation board, show that it can process 60.46 fps for $1360{\times}800$ resolution at 100MHz clock frequency.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.2 no.1
• /
• pp.32-44
• /
• 1999

Texture features can be incorporated in classification procedure to resolve class confusions. However, there have been few application-oriented studies made to evaluate the relative powers of texture analysis methods in a particular environment. This study evaluates the increases in the land-cover classification accuracy of the SPOT HRV multispectral data of Pusan Metropolitan area from texture processing. Twenty-four texture measures were derived from the SPOT HRV band 3 image. Each of these features were used in combination with the three spectral images in the classification of 10 land-cover classes. Supervised training and a Gaussian maximum likelihood classifier were used in the classification. It was found that while entropy produces the best empirical results in terms of the overall classification, other texture features can also largely improve the classification accuracies obtained by the use of the spectral images only. With the inclusion of texture, the classification for each category improves. Specially, urban built-up areas had much increase in accuracy. The results indicate that texture size 5 by 5 and 7 by 7 may be suitable at land cover classification of Pusan Metropolitan area.

• Journal of the Korean Society for Library and Information Science
• /
• v.43 no.3
• /
• pp.163-180
• /
• 2009

In this paper, we describe a developmental system for establishing personal information tendency based on user queries. For each query, the system classified it based on the category information using a kNN classifier. As category information, we used DDC field which is already assigned to each record in the database. The system accumulates category information for all user queries and the user's personalized feature for the target database. We then developed a personalized retrieval system reflecting the personalized feature to produce search result. Our system re-ranks the result documents by adding more weights to the documents for which categories match with the user's personalized feature. By using user's tendency information, the ambiguity problem of the word could be solved. In this paper, we conducted experiments for personalized search and word sense disambiguation (WSD) on a collection of Korean journal articles of science and technology arena. Our experimental result and user's evaluation show that the performance of the personalized search system and WSD is proved to be useful for actual field services.

• Journal of the Korea Convergence Society
• /
• v.9 no.11
• /
• pp.15-21
• /
• 2018

In this study, we try to detect anomalies on the network intrusion detection system by learning only one class. We use KDD CUP 1999 dataset, an intrusion detection dataset, which is used to evaluate classification performance. One class classification is one of unsupervised learning methods that classifies attack class by learning only normal class. When using unsupervised learning, it difficult to achieve relatively high classification efficiency because it does not use negative instances for learning. However, unsupervised learning has the advantage for classifying unlabeled data. In this study, we use one class classifiers based on support vector machines and density estimation to detect new unknown attacks. The test using the classifier based on density estimation has shown relatively better performance and has a detection rate of about 96% while maintaining a low FPR for the new attacks.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.3
• /
• pp.384-390
• /
• 2020

The reverberation effect on the sound when producing movies or VR contents is a very important factor in the realism and liveliness. The reverberation time depending the space is recommended in a standard called RT60(Reverberation Time 60 dB). In this paper, we propose a scene recognition technique for automatic reverberation editing. To this end, we devised a classification model that independently trains color images and predicted depth images in the same model. Indoor scene classification is limited only by training color information because of the similarity of internal structure. Deep learning based depth information extraction technology is used to use spatial depth information. Based on RT60, 10 scene classes were constructed and model training and evaluation were conducted. Finally, the proposed SCR + DNet (Scene Classification for Reverb + Depth Net) classifier achieves higher performance than conventional CNN classifiers with 92.4% accuracy.

• Journal of Biomedical Engineering Research
• /
• v.42 no.6
• /
• pp.287-294
• /
• 2021

Cardiotoxicity assessment of all drugs has been performed according to the ICH guidelines since 2005. Non-clinical evaluation S7B has focused on the hERG assay, which has a low specificity problem. The comprehensive in vitro proarrhythmia assay (CiPA) project was initiated to correct this problem, which presented a model for classifying the Torsade de pointes (TdP)-induced risk of drugs as biomarkers calculated through an in silico ventricular model. In this study, we propose a TdP-induced risk group classifier of artificial neural network (ANN)-based. The model was trained with 12 drugs and tested with 16 drugs. The ANN model was performed according to nine features, seven features, five features as an individual ANN model input, and the model with the highest performance was selected and compared with the classification performance of the qNet input logistic regression model. When the five features model was used, the results were AUC 0.93 in the high-risk group, AUC 0.73 in the intermediate-risk group, and 0.92 in the low-risk group. The model's performance using qNet was lower than the ANN model in the high-risk group by 17.6% and in the low-risk group by 29.5%. This study was able to express performance in the three risk groups, and it is a model that solved the problem of low specificity, which is the problem of hERG assay.

• International Journal of Computer Science & Network Security
• /
• v.22 no.7
• /
• pp.91-102
• /
• 2022

With the evolution of wireless sensor network (WSN) technology, the routing policy has foremost importance in the Internet of Things (IoT). A systematic routing policy is one of the primary mechanics to make certain the precise and robust transmission of wireless sensor networks in an energy-efficient manner. In an IoT environment, WSN is utilized for controlling services concerning data like, data gathering, sensing and transmission. With the advantages of IoT potentialities, the traditional routing in a WSN are augmented with decision-making in an energy efficient manner to concur finer optimization. In this paper, we study how to combine IoT-based deep learning classifier with routing called, Kriging Regressive Deep Belief Neural Learning (KR-DBNL) to propose an efficient data packet routing to cope with scalability issues and therefore ensure robust data packet transmission. The KR-DBNL method includes four layers, namely input layer, two hidden layers and one output layer for performing data transmission between source and destination sensor node. Initially, the KR-DBNL method acquires the patient data from different location. Followed by which, the input layer transmits sensor nodes to first hidden layer where analysis of energy consumption, bandwidth consumption and light intensity are made using kriging regression function to perform classification. According to classified results, sensor nodes are classified into higher performance and lower performance sensor nodes. The higher performance sensor nodes are then transmitted to second hidden layer. Here high performance sensor nodes neighbouring sensor with higher signal strength and frequency are selected and sent to the output layer where the actual data packet transmission is performed. Experimental evaluation is carried out on factors such as energy consumption, packet delivery ratio, packet loss rate and end-to-end delay with respect to number of patient data packets and sensor nodes.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.153-172
• /
• 2022

In this study, the authors intend to analyze factors contributing to research performance using Backpropagation Neural Network and Support Vector Machine. The analyzing factors contributing to lecturer research performance start from defining the features. The next stage is to collect datasets based on defining features. Then transform the raw dataset into data ready to be processed. After the data is transformed, the next stage is the selection of features. Before the selection of features, the target feature is determined, namely research performance. The selection of features consists of Chi-Square selection (U), and Pearson correlation coefficient (CM). The selection of features produces eight factors contributing to lecturer research performance are Scientific Papers (U: 154.38, CM: 0.79), Number of Citation (U: 95.86, CM: 0.70), Conference (U: 68.67, CM: 0.57), Grade (U: 10.13, CM: 0.29), Grant (U: 35.40, CM: 0.36), IPR (U: 19.81, CM: 0.27), Qualification (U: 2.57, CM: 0.26), and Grant Awardee (U: 2.66, CM: 0.26). To analyze the factors, two data mining classifiers were involved, Backpropagation Neural Networks (BPNN) and Support Vector Machine (SVM). Evaluation of the data mining classifier with an accuracy score for BPNN of 95 percent, and SVM of 92 percent. The essence of this analysis is not to find the highest accuracy score, but rather whether the factors can pass the test phase with the expected results. The findings of this study reveal the factors that have a significant impact on research performance and vice versa.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.6
• /
• pp.113-119
• /
• 2022

In software fault prediction, a multi classification model that predicts the fault severity category of a module can be much more useful than a binary classification model that simply predicts the presence or absence of faults. A small number of severity-based fault prediction models have been proposed, but no classifier using deep learning techniques has been proposed. In this paper, we construct MLP models with 3 or 5 hidden layers, and they have a structure with a fixed or variable number of hidden layer nodes. As a result of the model evaluation experiment, MLP-based deep learning models shows significantly better performance in both Accuracy and AUC than MLPs, which showed the best performance among models that did not use deep learning. In particular, the model structure with 3 hidden layers, 32 batch size, and 64 nodes shows the best performance.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.10
• /
• pp.2627-2642
• /
• 2023

Parkinson's disease (PD) is a typical, chronic neurodegenerative disease involving the concentration of dopamine, which can disrupt motor activity and cause different degrees of gait disturbance relevant to PD severity in patients. As current clinical PD diagnosis is a complex, time-consuming, and challenging task that relays on physicians' subjective evaluation of visual observations, gait disturbance has been extensively explored to make automatic detection of PD diagnosis and severity rating and provides auxiliary information for physicians' decisions using gait data from various acquisition devices. Among them, wearable sensors have the advantage of flexibility since they do not limit the wearers' activity sphere in this application scenario. In this paper, an attention-based temporal network (ATN) is designed for the time series structure of gait data (vertical ground reaction force signals) from foot sensor systems, to learn the discriminative differences related to PD severity levels hidden in sequential data. The structure of the proposed method is illuminated by Transformer Network for its success in excavating temporal information, containing three modules: a preprocessing module to map intra-moment features, a feature extractor computing complicated gait characteristic of the whole signal sequence in the temporal dimension, and a classifier for the final decision-making about PD severity assessment. The experiment is conducted on the public dataset PDgait of VGRF signals to verify the proposed model's validity and show promising classification performance compared with several existing methods.