• Journal of Communications and Networks
• /
• 제17권6호
• /
• pp.568-581
• /
• 2015

Predicting locations of users with portable devices such as IP phones, smart-phones, iPads and iPods in public wireless local area networks (WLANs) plays a crucial role in location management and network resource allocation. Many techniques in machine learning and data mining, such as sequential pattern mining and clustering, have been widely used. However, these approaches have two deficiencies. First, because they are based on profiles of individual mobility behaviors, a sequential pattern technique may fail to predict new users or users with movement on novel paths. Second, using similar mobility behaviors in a cluster for predicting the movement of users may cause significant degradation in accuracy owing to indistinguishable regular movement and random movement. In this paper, we propose a novel fusion technique that utilizes mobility rules discovered from multiple similar users by combining clustering and sequential pattern mining. The proposed technique with two algorithms, named the clustering-based-sequential-pattern-mining (CSPM) and sequential-pattern-mining-based-clustering (SPMC), can deal with the lack of information in a personal profile and avoid some noise due to random movements by users. Experimental results show that our approach outperforms existing approaches in terms of efficiency and prediction accuracy.

• 제어로봇시스템학회논문지
• /
• 제21권11호
• /
• pp.1008-1016
• /
• 2015

This paper proposes a method for detecting the front side of vehicles. The method can find the car side with a license plate even with complicated and cluttered backgrounds. A convolutional neural network (CNN) is used to solve the detection problem as a unified framework combining feature detection, classification, searching, and localization estimation and improve the reliability of the system with simplicity of usage. The proposed CNN structure avoids sliding window search to find the locations of vehicles and reduces the computing time to achieve real-time processing. Multiple responses of the network for vehicle position are further processed by a weighted clustering and probabilistic threshold decision method. Experiments using real images in parking lots show the reliability of the method.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2018년도 학술발표회
• /
• pp.29-29
• /
• 2018

전 세계적으로 홍수, 태풍, 폭설 등 기상이변에 따른 자연재난이 빈번히 발생하고 있으며, 국내의 경우 연간 약 5천억원 이상의 피해가 발생하고 있다. 미국 및 일본 등의 방재 선진국의 경우 재난 발생 전에 대비하는 재난관리가 중심을 이루고 있으며, 국내에서도 피해가 발생하기 전에 신속하게 재난피해를 예측 및 대비한다면 인명과 재산피해를 최소화 할 수 있을 것이라 판단된다. 따라서 본 연구에서는 신속하게 재난 피해를 예측하기 위해 기존에 함수 개발시 활발하게 사용되었던 다중회귀분석과 최근 이슈가 되고 있는 머신러닝(기계학습)을 활용하여 호우로 인한 피해를 사전에 예측하는 함수를 개발하였다. 행정안전부에서 구축하고 있는 재해연보 자료를 종속변수로 활용하였고, 기상요소 및 사회 경제적 요소를 설명변수로 사용하였다. 본 연구에서 개발된 호우피해 예측함수를 이용하여 호우피해를 예측하고, 이를 기반으로 사전 대비 차원의 재난관리를 실시한다면 자연재난으로 인한 피해를 줄이는데 큰 도움이 될 것으로 판단된다.

• Genomics & Informatics
• /
• 제17권2호
• /
• pp.20.1-20.5
• /
• 2019

Entity normalization, or entity linking in the general domain, is an information extraction task that aims to annotate/bind multiple words/expressions in raw text with semantic references, such as concepts of an ontology. An ontology consists minimally of a formally organized vocabulary or hierarchy of terms, which captures knowledge of a domain. Presently, machine-learning methods, often coupled with distributional representations, achieve good performance. However, these require large training datasets, which are not always available, especially for tasks in specialized domains. CONTES (CONcept-TErm System) is a supervised method that addresses entity normalization with ontology concepts using small training datasets. CONTES has some limitations, such as it does not scale well with very large ontologies, it tends to overgeneralize predictions, and it lacks valid representations for the out-of-vocabulary words. Here, we propose to assess different methods to reduce the dimensionality in the representation of the ontology. We also propose to calibrate parameters in order to make the predictions more accurate, and to address the problem of out-of-vocabulary words, with a specific method.

• 대한건축학회논문집:계획계
• /
• 제36권4호
• /
• pp.103-111
• /
• 2020

Perforated metal sheets are used as panels of facades for controlling environmental factors while ensuring user's visibility. Despite their functional potentials, only a specific direction of facades or an orientation of a building was considered in the relevant studies. This study proposed a design methodology for the perforated panel facades that reflects the location on the facades and the user's requirements. The optimization of quantitative and qualitative performance is achieved through communication between designers and users in a VR system. In optimizing quantitative performances, designers use machine learning techniques such as clustering and genetic algorithm to allocate optimal panels on the facades. In optimizing qualitative performances, through the VR system, users intervene in evaluating performances whose preferences are depending on them. The experiment using the office project showed that designers were able to make decisions based on clustering using GMM to optimize multiple quantitative performances. The gap between the target and final performance could be narrowed by limiting the types of perforated panels considering mass customization. In assessing visibility as a qualitative performance, users were able to participate in the design process using the VR system.

• 전기전자학회논문지
• /
• 제25권1호
• /
• pp.25-31
• /
• 2021

일반적인 신용카드 번호 인식 시스템은 정해진 위치에 카드를 배치했을 때에만 올바르게 동작하도록 설계되어 있다. 본 논문은, 저시력 장애인을 포함한 시력 취약 계층에게 보다 쉬운 사용자 경험을 제공하기 위해, 신용카드 내 16자리 숫자의 종횡비 특징을 이용한 자동 번호 인식 알고리즘을 제안한다. 제안하는 알고리즘은 형태학 연산을 통해 종횡비가 4:1 이상인 이미지 후보군을 찾고, 각각의 후보에 OCR과 BIN 번호 매칭 기술을 적용하여 신용카드 번호를 획득한다. OpenCV 및 Firebase ML에 기반한 실험 결과, 카드를 정해진 위치에 두지 않아도 77.75% 정확도로 카드 번호를 인식하였다.

• Journal of Information Processing Systems
• /
• 제17권3호
• /
• pp.556-570
• /
• 2021

The existing video expression recognition methods mainly focus on the spatial feature extraction of video expression images, but tend to ignore the dynamic features of video sequences. To solve this problem, a multi-mode convolution neural network method is proposed to effectively improve the performance of facial expression recognition in video. Firstly, OpenFace 2.0 is used to detect face images in video, and two deep convolution neural networks are used to extract spatiotemporal expression features. Furthermore, spatial convolution neural network is used to extract the spatial information features of each static expression image, and the dynamic information feature is extracted from the optical flow information of multiple expression images based on temporal convolution neural network. Then, the spatiotemporal features learned by the two deep convolution neural networks are fused by multiplication. Finally, the fused features are input into support vector machine to realize the facial expression classification. Experimental results show that the recognition accuracy of the proposed method can reach 64.57% and 60.89%, respectively on RML and Baum-ls datasets. It is better than that of other contrast methods.

• 스마트미디어저널
• /
• 제10권1호
• /
• pp.79-87
• /
• 2021

In recent years research on automatic font generation with machine learning mainly focus on using transformation-based methods, in comparison, generative model-based methods of font generation have received less attention. Transformation-based methods learn a mapping of the transformations from an existing input to a target. This makes them ambiguous because in some cases a single input reference may correspond to multiple possible outputs. In this work, we focus on font generation using the generative model-based methods which learn the buildup of the characters from noise-to-image. We propose a novel way to train a conditional generative deep neural model so that we can achieve font style control on the generated font images. Our research demonstrates how to generate new font images conditioned on both character class labels and character style labels when using the generative model-based methods. We achieve this by introducing a modified generator network which is given inputs noise, character class, and style, which help us to calculate losses separately for the character class labels and character style labels. We show that adding the character style vector on top of the character class vector separately gives the model rich information about the font and enables us to explicitly specify not only the character class but also the character style that we want the model to generate.

• 품질경영학회지
• /
• 제50권3호
• /
• pp.305-332
• /
• 2022

Purpose: As the public interest in entrepreneurship has been highlighted and entrepreneurship policies have been generated, this study is to construct Entrepreneurship Ecosystem (EE) models which have a significant relationship to national entrepreneurship with quantitative analysis. It aims to provide implications to EE policymakers that which national components are effective in cultivating innovative entrepreneurship and validate its EE quality based on quantitative performance goals. Methods: This study utilizes secondary data, categorized under the PESTLE factor from credible international organizations (WB, UNDP, GEM, GEDI, and OECD) to determine significant factors in the quality of the entrepreneurial ecosystem. This paper uses the Multiple Linear Regression (MLR) analysis to select the significant variables contributing to entrepreneurship performance. Using the AUC-ROC performance evaluation method for machine learning MLR results, this paper evaluates the performance of EE models so that it can allow approving EE quality by predicting potential performance. Results: Among nine hypothesis models, MLR analysis examines that the number of the Unicorn company, Unicorn companies' economic value, and entrepreneurship measured as GEI can be reasonable dependent variables to indicate the performance derived from EE quality. Rather than government policies and regulations, the social, finance, technology, and economic variables are significant factors of EE quality determining its performance. By having high Area Under Curve values under AUC-ROC analysis, accepted MLR models are regarded as having high prediction accuracy. Conclusion: Superior EE contributes to the outstanding Unicorn companies, and improvement in macro-environmental components can enhance EE quality.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 춘계학술발표대회
• /
• pp.385-387
• /
• 2022

With the tremendous rise in popularity of the Internet and technological advancements, many news keeps generating every day from multiple sources. As a result, the information (News) on the network has been highly increasing. The critical problem is that the volume of articles or news content can be overloaded for the readers. Therefore, the people interested in reading news might find it difficult to decide which content they should choose. Recommendation systems have been known as filtering systems that assist people and give a list of suggestions based on their preferences. This paper studies a personalized news recommendation system to help users find the right, relevant content and suggest news that readers might be interested in. The proposed system aims to build a hybrid system that combines collaborative filtering with content-based filtering to make a system more effective and solve a cold-start problem. Twitter social media data will analyze and build a user's profile. Based on users' tweets, we can know users' interests and recommend personalized news articles that users would share on Twitter.