• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.12
• /
• pp.4489-4501
• /
• 2014

We propose a novel lane detection method based on classification in image sequences. Both structural and statistical features of the extracted bright shape are applied to the neural network for finding correct lane marks. The features used in this paper are shown to have strong discriminating power to locate correct traffic lanes. The traffic lanes detected in the current frame is also used to estimate the traffic lane if the lane detection fails in the next frame. The proposed method is fast enough to apply for real-time systems; the average processing time is less than 2msec. Also the scheme of the local illumination compensation allows robust lane detection at nighttime. Therefore, this method can be widely used in intelligence transportation systems such as driver assistance, lane change assistance, lane departure warning and autonomous vehicles.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.4
• /
• pp.9-16
• /
• 2018

In this paper, we propose wide range of categorizes Artificial Intelligence technology as Learning, Inference, and Cognitive. Also, it analyzes 758 cases of open patents. For an analysis, target technologies were selected and categorized into specific areas to collect information about the patents. After removing noise, the patent information for each technology such as patent assignees and IPC code, was analyzed to evaluate the maturity of technology, the way ahead for research and development and the trends in core technology. This research presents directions of Artificial intelligence technology research and trend analysis of core Artificial Intelligent technology using quantitative analysis of patent information. Also Artificial intelligence technology requires technological development necessity through close cooperation in diverse fields.

• Journal of Intelligence and Information Systems
• /
• v.3 no.2
• /
• pp.83-91
• /
• 1997

DELVAUX is an inductive learning environment that learns Bayesian classification rules from a set o examples. In DELVAUX, a genetic a, pp.oach is employed to learn the best rule-set, in which a population consists of rule-sets and rule-sets generate offspring by exchanging some of their rules. We have explored a meta-learning a, pp.oach in the DELVAUX learning environment to improve the classification performance of the DELVAUX system. The meta-learning a, pp.oach learns the bias of a classifier so that it can evaluate the prediction made by the classifier for a given example and thereby improve the overall performance of a classifier system. The paper discusses the meta-learning a, pp.oach in details and presents some empirical results that show the improvement we can achieve with the meta-learning a, pp.oach.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2003.05a
• /
• pp.206-213
• /
• 2003

Application of existing decision tree algorithms for Internet retail customer classification is apt to construct a bushy tree due to imprecise source data. Even excessive analysis may not guarantee the effectiveness of the business although the results are derived from fully detailed segments. Thus, it is necessary to determine the appropriate number of segments with a certain level of abstraction. In this study, we developed a stopping rule that considers the total amount of information gained while generating a rule tree. In addition to forwarding from root to intermediate nodes with a certain level of abstraction, the decision tree is investigated by the backtracking pruning method with misclassification loss information.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2005.11a
• /
• pp.295-303
• /
• 2005

Semantic Web and its related technologies have been opening the era of information sharing via Web. In the meantime, there are several huddles to overcome toward the new era and one of the major huddles is information integration issue unless we build and use a single unified but huge ontology which address everything in the world. Particularly in e-business area, information integration problem must be a great concern in search and comparison of products from various internet shopping sites and e-marketplaces. To overcome such an information integration problem, we propose an ontology driven mapping algorithm between heterogeneous product classification and description frameworks. We also perform comparative evaluation of the proposed mapping algorithm against a well-known ontology mapping tool, PROMPT.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2003.05a
• /
• pp.435-442
• /
• 2003

정보 추출(Information Retrieval) 시스템은 거대한 양의 정보들 가운데 필요한 정보의 적절한 탐색을 도와주기 위한 도구이다. 이는 사용자가 요구하는 정보를 보다 정확하고 보다 효과적이면서 보다 효율적으로 전달해주어야만 한다. 그러기 위해서는 문서내의 무수히 많은 속성들 가운데 해당 문서의 특성을 잘 반영하는 속성만을 선별해서 적절히 활용하는 것이 절실히 요구된다. 이에 본 연구는 기존의 한글 문서 분류시스템(CB_TFIDF)[1]의 정확도와 신속성 두 가지 측면의 성능향상에 초점을 두고 있다. 기존의 영문 텍스트 문서 분류시스템에 적용되었던 다양한 속성선택 기법들 가운데 잘 알려진 세가지 즉, Information Gain, Odds Ratio, Document Frequency Thresholding을 통해 선별적인 사례베이스를 구성한 다음에 한글 텍스트 문서 분류시스템에 적용시켜서 성능을 비교 평가한 후, 한글 문서 분류시스템에 가장 적절한 속성선택 기법과 속성 선택에 대한 가이드라인을 제시하고자 한다.

• International Journal of Advanced Culture Technology
• /
• v.9 no.3
• /
• pp.291-297
• /
• 2021

Recently, as the number of new and variant malicious codes has increased exponentially, malware warnings are being issued to PC and smartphone users. Malware is becoming more and more intelligent. Efforts to protect personal information are becoming more and more important as social issues are used to stimulate the interest of PC users and allow users to directly download malicious codes. In this way, it is difficult to prevent malicious code because malicious code infiltrates in various forms. As a countermeasure to solve these problems, many studies are being conducted to apply deep learning. In this paper, we investigate and analyze various deep learning methods to detect and classify malware.

• Journal of Korea Multimedia Society
• /
• v.24 no.5
• /
• pp.651-658
• /
• 2021

In this paper, we propose methods for character recognition and classification for table items through deep learning. First, table areas are detected in a document image through CNN. After that, table areas are separated by separators such as vertical lines. The text in document is recognized through a neural network combined with CNN and RNN. To correct errors in the character recognition, multiple candidates for the recognized result are provided for a sentence which has low recognition accuracy.

• International Journal of Computer Science & Network Security
• /
• v.22 no.4
• /
• pp.374-386
• /
• 2022

Nowadays, research in deep learning leveraged automated computing and networking paradigm evidenced rapid contributions in terms of Software Defined Networking (SDN) and its diverse security applications while handling cybercrimes. SDN plays a vital role in sniffing information related to network usage in large-scale data centers that simultaneously support an improved algorithm design for automated detection of network intrusions. Despite its security protocols, SDN is considered contradictory towards DDoS attacks (Distributed Denial of Service). Several research studies developed machine learning-based network intrusion detection systems addressing detection and mitigation of DDoS attacks in SDN-based networks due to dynamic changes in various features and behavioral patterns. Addressing this problem, this research study focuses on effectively designing a multistage hybrid and intelligent deep learning classifier based on modified deep forest classification to detect DDoS attacks in SDN networks. Experimental results depict that the performance accuracy of the proposed classifier is improved when evaluated with standard parameters.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.473-477
• /
• 2001

Artificial neural network (ANN) is known to identify relationships even when some of the input data are very complex, ill-defined and ill-structured. One of the advantages in ANN is that it can discriminate the linearly inseparable data. This study presents an application of ANN to classify and predict the symptomatic status of HIV/AIDS patients. Even though ANN techniques have been applied to a variety of areas, this study has a substantial contribution to the HIV/AIDS care and prevention planning area. ANN model in classifying both the HIV and AIDS status of HIV/AIDS patients is developed and analyzed. The diagnostic accuracy of the ANN in classifying both the HIV status and AIDS status of HIV/AIDS status is evaluated. Several different ANN topologies are applied to AIDS Cost and Services Utilization Survey (ACSUS) datasets in order to demonstrate the model\`s capability. If ANN design models are different, it would be interesting to see what influence would have on classification of HIV/AIDS-related persons.