• 한국빅데이터학회지
• /
• 제7권1호
• /
• pp.49-61
• /
• 2022

스포츠 분야에서는 팀 전략 구상과 마케팅 등 팀 운영에 있어서, 데이터 분석의 비중이 점점 더 커지고 있다. 특히, 한국프로야구에서는 한 시즌이 끝나면 FA, 트레이드 등 다음 해 팀 전략을 구상하기 위해서 선수 영입과 선수 육성 등의 계획을 수립하는데, 이 때 선수들의 다음 해 성적을 예측하는 것이 매우 중요하다. 본 연구에서는 타자만으로 대상을 한정지어 다음 해의 성적이 상승할지를 예측해보고자 하였다. 상승 및 하락의 기준이 되는 기록으로는, 계산하기 쉽고 팀 득점과의 관계가 높은 OPS로 하였다. 본 연구에서 데이터는 한국프로야구 1982년부터 2021년까지 40년간의 정규시즌 데이터를 사용하였고, 실험 방법으로는 11개의 머신러닝 분류 모델을 사용하였다. OPS의 상승 및 하락 여부를 예측해본 결과, RBF SVM, Neural Net, Gaussian Process, AdaBoost가 다른 분류 모델에 비해 정확도가 높게 나왔고 나이는 정확도에 큰 영향을 주지 못했다.

• 음성과학
• /
• 제15권2호
• /
• pp.55-65
• /
• 2008

Recently, the multi-model based speech recognizer has been used quite successfully for noisy speech recognition. For the selection of the reference HMM (hidden Markov model) which best matches the noise type and SNR (signal to noise ratio) of the input testing speech, the estimation of the SNR value using the VAD (voice activity detection) algorithm and the classification of the noise type based on the GMM (Gaussian mixture model) have been done separately in the multi-model framework. As the SNR estimation process is vulnerable to errors, we propose an efficient method which can classify simultaneously the SNR values and noise types. The KL (Kullback-Leibler) distance between the single Gaussian distributions for the noise signal during the training and testing is utilized for the classification. The recognition experiments have been done on the Aurora 2 database showing the usefulness of the model compensation method in the multi-model based speech recognizer. We could also see that further performance improvement was achievable by combining the probability density function of the MCT (multi-condition training) with that of the reference HMM compensated by the D-JA (data-driven Jacobian adaptation) in the multi-model based speech recognizer.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제16권1호
• /
• pp.44-51
• /
• 2016

Recently, artificial intelligence has reached the level of top information technologies that will have significant influence over many aspects of our future lifestyles. In particular, in the fields of machine learning technologies for classification and decision-making, there have been a lot of research efforts for solving estimation and control problems that appear in the various kinds of portfolio management problems via data-driven approaches. Note that these modern data-driven approaches, which try to find solutions to the problems based on relevant empirical data rather than mathematical analyses, are useful particularly in practical application domains. In this paper, we consider some applications of modern data-driven machine learning methods for portfolio management problems. More precisely, we apply a simplified version of the sparse Gaussian process (GP) classification method for classifying users' sensitivity with respect to financial risk, and then present two portfolio management issues in which the GP application results can be useful. Experimental results show that the GP applications work well in handling simulated data sets.

• Journal of Mechanical Science and Technology
• /
• 제18권11호
• /
• pp.1923-1931
• /
• 2004

Manufacturing of complex surface plates in stern and stem is a major factor in cost of a preliminary ship design by computing process. If these hull plate parts are effectively classified, it helps to compute the processing cost and find the way to cut-down the processing cost. This paper presents a new method to classify surface plates effectively in the preliminary ship design using neural network. A neural-network-based ship hull plate classification program was developed and tested for the automatic classification of ship design. The input variables are regarded as Gaussian curvature distributions on the plate. Various applicable rules of network topology are applied in the ship design. In automation of hull plate classification, two different numbers of input variables are used. By observing the results of the proposed method, the effectiveness of the proposed method is discussed. As a result, high prediction rate was achieved in the ship design. Accordingly, to the initial design stage, the ship hull plate classification program can be used to predict the ship production cost. And the proposed method will contribute to reduce the production cost of ship.

• 대한원격탐사학회지
• /
• 제20권6호
• /
• pp.419-425
• /
• 2004

본 연구에서는 Dempster-Shafer evidence theory에 기반하여 Gaussian 질량 함수를 사용하는 융합 기법을 제안하고 있다. Dempster-Shafer 융합은 비정확성과 불확실성 measures를 각각 belief 함수와 plausibility 함수로 나타내며 이 두 함수 값 사이의 간격을 나타내는 "belief interval"에 의해 불확실성의 정도가 표현된다. 이러한 Dempster-Shafer 융합기술을 이용하여 서로 다른 센서에서 수집된 영상 자료를 융합하여 사용하여 분류 결과의 정확성을 높이고 특히 분류를 위한 매개변수를 추정하는 훈련과정에서 복합 클래스를 설정할 수 있어 단순 클래스 설정으로 인한 훈련과정이 어려움을 피할 수 있다. 이 연구에서는 경기도 용인/능평 지역에서 관측 된 KOMPSAT EOC의 범색 영상 자료와 LANDSAT ETM+의 식생지수 자료에 대해 제안된 Dempster-Shafer 융합기술을 이용하여 분류 실험을 수행하였고 분류 결과는 서로 다른 센서간의 영상자료 융합을 위한 제안된 기법의 잠재적 효과성을 보여주고 있다.

• 대한조선학회논문집
• /
• 제46권6호
• /
• pp.675-687
• /
• 2009

In general, the forming process of the curved hull plates consists of sub tasks, such as roll bending, line heating, and triangle heating. In order to complement the automated curved hull forming system, it is necessary to develop an algorithm to classify the curved hull plates of a ship into standard shapes with respect to the techniques of forming task, such as the roll bending, the line heating, and the triangle heating. In this paper, the curved hull plates are classified by four standard shapes and the combination of them, or saddle, convex, flat, cylindrical shape, and the combination of them, that are related to the forming tasks necessary to form the shapes. In preprocessing, the Gaussian curvature and the mean curvature at the mid-point of a mesh of modeling surface by Coon's patch are calculated. Then the nearest neighbor method to classify the input plate type is applied. Tests to verify the developed algorithm with sample plates of a real ship data have been performed.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2022년도 춘계학술대회
• /
• pp.437-440
• /
• 2022

현대 사회는 4차 산업 혁명과 IoT 기술 등의 발전으로 영상 처리 분야의 활용이 급증하고 있다. 특히, 에지 검출은 이미지 분류, 객체 검출 등 영상 처리 응용에서 필수적인 전처리 과정으로 여러 분야에서 널리 사용되고 있다. 에지를 검출하기 위한 기존의 방법에는 소벨 필터(Sobel edge detection filter), 로버츠 필터(Roberts edge detection filter), 프리윗 필터(Prewitt edge detection filter), LoG(Laplacian of Gaussian) 등이 있다. 하지만 기존의 방법들은 명암도가 낮은 저조도 환경에서 에지 검출 특성이 다소 미흡한 성능을 보인다는 단점이 있다. 따라서 본 논문에서는 저조도 환경에서도 에지 검출 특성을 높이기 위해 명암도 분석에 기반한 에지 검출 알고리즘을 제안한다.

• Journal of Multimedia Information System
• /
• 제2권4호
• /
• pp.303-310
• /
• 2015

Vulnerable road users such as bike, motorcycle, small automobiles, and etc. are easily attacked or threatened with bigger vehicles than them. So this paper suggests a new approach two-wheelers detection system riding on people based on modified histogram of oriented gradients (HOGs) which is weighted by curvature and local correlation coefficient. This correlation coefficient between two variables, in which one is the person riding a bike and other is its background, can represent correlation relation. First, we extract edge vectors using the curvature of Gaussian and Histogram of Oriented Gradients (HOG) which includes gradient information and differential magnitude as cell based. And then, the value, which is calculated by the correlation coefficient between the area of each cell and one of bike, can be used as the weighting factor in process for normalizing the HOG cell. This paper applied the Adaboost algorithm to make a strong classification from weak classification. The experimental results validate the effectiveness of our proposed algorithm show higher than that of the traditional method and under challenging, such as various two-wheeler postures, complex background, and even conclusion.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2004년도 ICEIC The International Conference on Electronics Informations and Communications
• /
• pp.237-243
• /
• 2004

A multichannel smart sound sensor capable to detect and identify sound events in noisy conditions is presented in this paper. Sound information extraction is a complex task and the main difficulty consists is the extraction of high­level information from an one-dimensional signal. The input of smart sound sensor is composed of data collected by 5 microphones and its output data is sent through a network. For a real time working purpose, the sound analysis is divided in three steps: sound event detection for each sound channel, fusion between simultaneously events and sound identification. The event detection module find impulsive signals in the noise and extracts them from the signal flow. Our smart sensor must be capable to identify impulsive signals but also speech presence too, in a noisy environment. The classification module is launched in a parallel task on the channel chosen by data fusion process. It looks to identify the event sound between seven predefined sound classes and uses a Gaussian Mixture Model (GMM) method. Mel Frequency Cepstral Coefficients are used in combination with new ones like zero crossing rate, centroid and roll-off point. This smart sound sensor is a part of a medical telemonitoring project with the aim of detecting serious accidents.

• 한국멀티미디어학회논문지
• /
• 제14권12호
• /
• pp.1544-1548
• /
• 2011

A process of choosing a subset of original features, so called feature selection, is considered as a crucial preprocessing step to image processing applications. There are already large pools of techniques developed for machine learning and data mining fields. In this paper, basically two methods, non-feature selection and feature selection, are investigated to compare their predictive effectiveness of classification. Color co-occurrence feature is used for defining image features. Standard Sequential Forward Selection algorithm are used for feature selection to identify relevant features and redundancy among relevant features. Four color spaces, RGB, YCbCr, HSV, and Gaussian space are considered for computing color co-occurrence features. Gray-level image feature is also considered for the performance comparison reasons. The experimental results are presented.