• Annual Conference on Human and Language Technology
• /
• 1999.10e
• /
• pp.191-195
• /
• 1999

자연스럽고 명료한 한국어 Text-to-Speech 변환 시스템을 위해서 음소의 지속 시간을 제어하는 일은 매우 중요하다. 음소의 지속 시간은 여러 가지 문맥 정보에 의해서 변화하므로 제어 규칙에 의존하기 보다 방대한 데이터베이스를 이용하여 통계적인 기법으로 음소의 지속 시간에 변화를 주는 요인을 찾아내려고 하는 것이 지금의 추세이다. 본 연구에서도 트리기반 모델링 방법중의 하나인 CART(classification and regression tree) 방법을 사용하여 회귀 트리를 생성하고, 생성된 트리에 기반하여 음소의 지속 시간 예측 모델과, 자연스러운 끊어 읽기를 위한 휴지 기간 예측 모델을 제안하고 있다. 실험에 사용한 음성코퍼스는 550개의 문장으로 구성되어 있으며, 이 중 428개 문장으로 회귀 트리를 학습시켰고, 나머지 122개의 문장으로 실험하였다. 모델의 평가를 위해서 실제값과 예측값과의 상관관계를 구하였더니 음소의 지속 시간을 예측하는 회귀 트리에서는 상관계수가 0.84로 계산되었고, 끊어 읽는 경계에서의 휴지 기간을 예측하는 회귀 트리에서는 상관계수가 0.63으로 나타났다.

• East Asian Journal of Business Economics (EAJBE)
• /
• v.2 no.2
• /
• pp.37-47
• /
• 2014

Purpose - In this study, it is attempted to examine the Islamic banking practice in Iran based on new scientific methods. Design/methodology/approach- It is used the financial ratios demonstrating healthy or non-healthy of banks to assess the financial health of listed banks in Tehran Stock Exchange. The assessment of these ratios with use of decision tree as a non-parametric method for modeling is recommended for presenting this model. Information about the financial health of banks could be effective on the decisions of different groups of banks' financial reports users, including shareholders, auditors, stock exchange, central bank and etc. Findings - the results of the study show that Decision Tree is strong approach in order to classifying Islamic banks in Iran. Originality/value- So far, several studies have been conducted in various countries on the topic of this study. Considering the importance Islamic banking, it is almost the first study in Iran and the outcomes of the study may helpful to Iranian economy.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.13 no.1
• /
• pp.1-8
• /
• 2010

This paper presents global and optimal solution for weapon assignment problems using the Munkres assignment algorithm. We propose a new modeling method of weapon assignment problems concerning some constraints of weapon systems. In this paper, we compares the Munkres weapon assignment algorithm with two other algorithms employing a search tree model in terms of computational complexity and performance. One is an optimal algorithm using exhausted search and the other is a greedy algorithm which selects the first search result as a solution. The experiment results show that the Munkres weapon assignment algorithm has better performance and less computational complexity in comparison with the two other algorithms.

• Speech Sciences
• /
• v.14 no.2
• /
• pp.91-103
• /
• 2007

The purpose of this paper is to develop a model enabling to predict the prosody of Korean text-to-speech synthesis using the CART and SKES algorithms. CART prefers a prediction variable in many instances. Therefore, a partition method by F-Test was applied to CART which had reduced the number of instances by grouping phonemes. Furthermore, the quality of the text-to-speech synthesis was evaluated after applying the SKES algorithm to the same data size. For the evaluation, MOS tests were performed on 30 men and women in their twenties. Results showed that the synthesized speech was improved in a more clear and natural manner by applying the SKES algorithm.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06c
• /
• pp.362-363
• /
• 2012

본 논문에서는 한 장의 실제 나무 영상이 주어졌을 시, 사실적인 3차원 나무 모델링(modeling) 및 자가생장(self-growth) 표현을 위한 방법을 소개하도록 한다. 스켈레톤기반의 간략화(skeleton-based abstraction)를 이용하여 동일한 나무 몸통(trunk)을 갖는 다양한 나무 모델생성과 함께 나무의 다면체구조(manifold structure)를 고려한 지오데식 커널(geodesic kernel)을 이용하여 나무의 자가생장을 표현한다. 나무의 자가생장은 사전 정의된 나무 굵기, 전체 크기, 그리고 가지증식 순서정보와 같은 상대적 성장정보(allometric information)를 동시 이용하여 상대적인 나무 생장(allometric tree growth)을 표현하도록한다. 한편, 보여지지않는 나무 가지와 잎들에 대해선, 나무구조는 로컬하게 자기유사성(local self-similarity)을 갖는다라는 고전적인 절차적(conventional procedural) 가정을 이용하여 자동적으로 생성토록한다. 실제영상을 이용한 몇몇들의 실험을 통해 보다 효과적으로 나무 모델 및 생장 표현이 가능함을 보여주도록한다.

• Smart Media Journal
• /
• v.13 no.4
• /
• pp.9-15
• /
• 2024

Tree-based algorithms have been the dominant methods used build a prediction model for tabular data. This also includes personal credit data. However, they are limited to compatibility with categorical and numerical data only, and also do not capture information of the relationship between other features. In this work, we proposed an ensemble model using the Transformer architecture that includes text features and harness the self-attention mechanism to tackle the feature relationships limitation. We describe a text formatter module, that converts the original tabular data into sentence data that is fed into FinBERT along with other text features. Furthermore, we employed FT-Transformer that train with the original tabular data. We evaluate this multi-modal approach with two popular tree-based algorithms known as, Random Forest and Extreme Gradient Boosting, XGBoost and TabTransformer. Our proposed method shows superior Default Recall, F1 score and AUC results across two public data sets. Our results are significant for financial institutions to reduce the risk of financial loss regarding defaulters.

• The KIPS Transactions:PartC
• /
• v.11C no.2
• /
• pp.183-192
• /
• 2004

As the use of the Internet has explosively increased, it is likely for the Internet to be exposed to various attacks. Modeling the Internet attacks is essential to simulate the attacks. However, the existing studies on attack modeling have mainly focused on classifying and categorizing the attacks and consequently they are not suitable to representing attack scenarios in the Internet security simulation. In this paper, we introduce the existing methods of attack modeling, and propose an adapted attack modeling to properly express the properties for the Internet security simulator. The adapted attack modeling suggests a solution to the problems of the existing attack tree modelings, such as difficulty of composing complex scenarios ambiguity of attack sequence, lack of system state information. And it can represent simultaneous, precise time-dependent attack, and attack period, which are nearly impossible to be represented in many other existing methods.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.7
• /
• pp.319-324
• /
• 2006

We Propose two new methods for non-keyword modeling to improve the performance of speaker- and vocabulary-independent keyword spotting system. The first method is decision tree clustering of monophone at the state level instead of monophone clustering method based on K-means algorithm. The second method is multi-state multiple mixture modeling at the syllable level rather than single state multiple mixture model for the non-keyword. To evaluate our method, we used the ETRI speech DB for training and keyword spotting test (closed test) . We also conduct an open test to spot 100 keywords with 400 sentences uttered by 4 speakers in an of fce environment. The experimental results showed that the decision tree-based state clustering method improve 28%/29% (closed/open test) than the monophone clustering method based K-means algorithm in keyword spotting. And multi-state non-keyword modeling at the syllable level improve 22%/2% (closed/open test) than single state model for the non-keyword. These results show that two proposed methods achieve the improvement of keyword spotting performance.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.3
• /
• pp.121-134
• /
• 2017

With rapid increasing the development and use of IoT Devices, requirements for safe IoT devices and services such as reliability, security are also increasing. In Security engineering, SDLC (Secure Development Life Cycle) is applied to make the trustworthy system. Secure Development Life Cycle has 4 big steps, Security requirements, Design, Implementation and Operation and each step has own goals and activities. Deriving security requirements, the first step of SDLC, must be accurate and objective because it affect the rest of the SDLC. For accurate and objective security requirements, Threat modeling is used. And the results of the threat modeling can satisfy the completeness of scope of analysis and the traceability of threats. In many countries, academic and IT company, a lot of researches about drawing security requirements systematically are being done. But in domestic, awareness and researches about deriving security requirements systematically are lacking. So in this paper, I described about method and process to drawing security requirements systematically by using threat modeling including DFD, STRIDE, Attack Library and Attack Tree. And also security requirements are described via Common Criteria for delivering objective meaning and broad use of them.

• Journal of Internet Computing and Services
• /
• v.10 no.5
• /
• pp.65-78
• /
• 2009

The most commonly adopted approach to find valuable information from tree data is to extract frequently occurring subtree patterns from them. Because mining frequent tree patterns has a wide range of applications such as xml mining, web usage mining, bioinformatics, and network multicast routing, many algorithms have been recently proposed to find the patterns. However, existing tree mining algorithms suffer from several serious pitfalls in finding frequent tree patterns from massive tree datasets. Some of the major problems are due to (1) modeling data as hierarchical tree structure, (2) the computationally high cost of the candidate maintenance, (3) the repetitious input dataset scans, and (4) the high memory dependency. These problems stem from that most of these algorithms are based on the well-known apriori algorithm and have used anti-monotone property for candidate generation and frequency counting in their algorithms. To solve the problems, we base a pattern-growth approach rather than the apriori approach, and choose to extract maximal frequent subtree patterns instead of frequent subtree patterns. The proposed method not only gets rid of the process for infrequent subtrees pruning, but also totally eliminates the problem of generating candidate subtrees. Hence, it significantly improves the whole mining process.