Earticle

본 연구는 2023년 충청남도 홍성군 산불피해지 대상으로 Maximum Likelihood Classification(MLC), Random Forest(RF), Support Vector Machine(SVM), K-Nearest Neighbors(KNN) 이상 4가지 머신러닝 알고리 즘 모델의 산불피해 강도 분류 성능을 파악하기 위해 실시되었다. 검증데이터와 비교한 결과, 모델 총 성능은 Random Forest(RF)의 Overall Accuacy(OA)와 Cohen’s Kappa 계수가 97.2%와 0.95로 가장 높은 성능을 보였다. RF는 여러 개의 의사결정 나무들을 결합하여 최종 예측을 하므로 데이터의 불균형을 해소할 수 있다. 또한, 산불피해 강도에 따른 성능은 모든 머신러닝 알고리즘 모델에서 열해 피해지가 수관화 및 지표화 피해지에 비해 낮은 탐지 성능을 나타냈다. 이는 수관화ㆍ열해 피해지의 분광 반사율 패턴이 유사하고, 열해·지표화 피해지의 경우 복잡한 수관층 구조로 인해 탐지 성능이 낮은 것으로 판단된다. 이와 같은 머신러닝 알고리즘 모델을 통한 산불피해 강도별 면적을 산출은 신속한 복구계획 수립뿐만 아니라 산불로 인한 온실가스 배출량 산정에 기초자료를 제공할 수 있을 것이다.

This study aims to evaluate the wildfire severity classification performance of four machine learning algorithm models, Maximum Likelihood Classification(MLC), Random Forest(RF), Support Vector Machine(SVM) and K-Nearest Neighbors(KNN), in the wildfire areas of Hongseong-gun, Chungcheongnam-do, in 2023. Compared with the verification dataset, the overall performance of the models indicated that the Random Forest(RF) showed the highest accuracy. The highest performance of the Random Forest(RF) model can be explained by its ability to combine multiple decision trees for a final prediction. Furthermore, the performance based on wildfire severity showed that all machine learning algorithm models had lower detection performance in heat damage areas compared with crown and surface fire areas. Crown and heat damage areas have similar spectral reflectance patterns, and heat damage and surface fire areas are considered to have shown lower performance due to the complex canopy structure. Machine learning based burn severity estimation provide critical baselines for rapid recovery planning and greenhouse gas emissions calculations.

In heterogeneous landscapes, high-resolution land cover classification is vital for planning, ecological monitoring, and green infrastructure management. This study evaluates the performance of five machine learning algorithms: Decision Tree (DT), Naïve Bayes, Support Vector Machine (SVM), Random Tree (RT), and K-Nearest Neighbors (KNN), using UAV multispectral imagery and object-based image analysis (OBIA). Five scenarios were designed to compare algorithm accuracy. Additionally, a sixth scenario applied the best-performing algorithm to a feature subset selected through Recursive Feature Elimination (RFE), to examine the effect of feature optimization. RT achieved the highest overall classification accuracy (76.75%) and Kappa coefficient (0.7066), while SVM showed limited performance in complex environments. Height features contributed most to accuracy improvements, followed by spectral and geometric features. In class-specific analysis, the Naïve Bayes algorithm yielded the highest Producer’s Accuracy (90.86%) for forest-type land cover but had a lower User’s Accuracy (70.41%), indicating overclassification. In contrast, RT showed more balanced performance (PA = 87.09%, UA = 85.71%), suggesting greater reliability. The results demonstrate the benefits of integrating algorithm selection with feature optimization to improve classification accuracy in complex settings. This approach provides methodological insights for fine-scale mapping of vegetated areas and supports future applications in landscape monitoring and urban green space assessment.

We may be given a set of observations with the classes or clusters. The aim of this article is to provide an up-to-date review of different approaches to classification, compare their performance on a wide range of challenging data-sets. In this paper, machine learning algorithm classifiers based on CART, C4.5, CAL5, FACT, QUEST and statistical discriminant analysis are compared on various datasets in classification error rate and algorithms.

인식 대상 학습 모델이 분류되어 있지 않거나 명확하게 분류되지 않은 경우 어휘 인식을 결정하지 못하여 인 식률이 저하되며 학습 모델 분류 형태가 변경되거나 새로운 학습 모델이 추가되면 인식 모델의 결정 트리 구조가 변 경되어야 하는 구조적 문제가 발생한다. 이러한 문제점을 해결하기 위하여 학습 모델 분류를 위한 결정 트리 학습 알 고리즘을 제안한다. 음운 현상이 충분히 반영된 음성 데이터베이스를 구성하고 학습 효과를 확보하기 위하여 학습 모 델 분류를 위한 결정 트리 방법을 사용하였다. 본 연구에서는 실내 환경에 대하여 어휘 종속 인식과 어휘 독립 인식 실험을 수행한 결과 실내 환경의 어휘 종속 실험에서는 98.3%의 인식 성능을 보였고, 어휘 독립 실험에서 98.4%의 인식 성능을 보였다.

Target learning model is not recognized in this category or not classified clearly failed to determine if the vocabulary recognition is reduced. Form of classification learning model is changed or a new learning model is added to the recognition decision tree structure of the model should be changed to a structural problem. In order to solve these problems, a decision tree learning model for classification learning algorithm is proposed. Phonological phenomenon reflected sound enough to configure the database to ensure learning a decision tree learning model for classifying method was used. In this study, the indoor environment-dependent recognition and vocabulary words for the experimental results independent recognition vocabulary of the indoor environment-dependent recognition performance of 98.3% in the experiment showed, vocabulary independent recognition performance of 98.4% in the experiment shown.

This paper reviews ordinal decision tree algorithms for ordinal classification, exploring theoretical foundations, key algorithms (MDT, QMDT), specialized splitting criteria (Ordinal Gini, Weighted Information Gain), and ensemble methods. It discusses applications in healthcare and social sciences, highlighting interpretability and flexibility while acknowledging overfitting and instability. As implications for future research, this study points out advantages such as interpretability and flexibility, and limitations such as overfitting and instability.

BI-RADS의 구분에 따라 Algorithm의 변화로 영상의 질을 개선하는 기법을 이용하여 시각적 가시화의 차이 를 일치도와 민감도로 평가하였다. 유방촬영을 시행한 172명을 대상으로 유방촬영의 판독소견과 자료 체계의 신뢰도 를 평가하였다. Category 5단계(C0,C1,C2,C3,C4), 유방 실질 함유량 4단계(Fatty, Fibroglandular, Heterogeneous nodular, Diffuse dense), 병변 3종류(석회화, 결절, 종괴)로 분류하여 TE와 PV의 신뢰도를 평가하고, 민감도와 진단의 일치도, 정확도를 평가함으로 다방면의 융복합적 분석을 하였다. TE보다 PV가 병변의 신뢰도와 민감도와 정확도가 높았다. 유방 실질 함유량에 따른 평가에서 정확도는 PV가 높았다. TE에서는 Fatty는 모두에서, Fibroglandular는 종 괴와 석회화가, Diffuse dense는 결절과 석회화가 구별이 용이하였다. PV에서는 Fatty는 모두에서, Fibroglandular는 결절, Heterogeneous nodular은 결절과 종괴, Diffuse dense는 결절과 석회화가 구별이 용이하였다. 민감도는 결절은 Fatty, Fibroglandular, Heterogeneous nodular에서 TE가 더 민감했고, 종괴에서는 Heterogeneous nodular, Diffuse dense 에서 TE가 더 민감했으며, 석회화에서는 모두에서 TE가 더 민감하였다. 이에 Algorithm 기법을 적절히 변화시켜 활 용한다면 진단과 판독에 정확성을 높일 것이라 사료된다.

Image availability evaluated by the degree of agreement and sensitive using the process improve visualization according to the Algorithm modification in Image Post-Processing. Reliability measured by the Breast Imaging Reporting and Data System. 172 patients visit same period divided by BI-RADS, category five stages, and contents of breast parenchyma into Calcification, Nodule and Mass. Evaluated the TE/PV image reliability, visualization sensitive, agreement of diagnosis. Convergence analysis was an in various fields. According to the result of this research, PV has higher sensitive and accuracy about lesions than TE visual and there is a difference insensitive by contents of breast parenchyma. Therefore, practical use of Algorithm Modification(Tissue Equalization: TE, Premium View: PV) is expected to improve more accurate, useful diagnosis, which has not been easy until now.

다양한 개발자들이 참여하는 환경에서 서로의 상호작용을 이해하고 시스템 개발부분을 공유하는 오픈 플랫폼 소프 트웨어를 성공적으로 개발하기 위해서는 일관성있는 시스템 및 소프트웨어 요구사항에 대한 관리가 필요하다. 본 논 문에서는 다양한 요구사항 문장을 체계적으로 관리하기 위하여 요구사항 문장 집합을 구문론적으로 유사한 집단으 로 군집화하며, 올바르게 작성되지 않은 요구사항 문장을 사용자가 수정할 수 있도록 이와 유사한 표준 템플릿을 자 동으로 추천하는 방안을 제안한다. 철도 상호호환 통합 플랫폼 설계 및 개발 도메인에서 올바르게 작성된 총 500개 의 요구사항 문장에 대한 템플릿 집합을 대상으로 요구사항 문장의 분류 및 추천 정확도를 측정하였다. 실험에서 요 구사항 문장의 군집화에 대한 평균 분류 정확도는 99% 이상을 나타냈으며, 유사한 문장의 평균 추천 정확도는 93% 이상을 나타내었다.

To successfully develop a software system for an open platform, which involves many software and hardware developers and interacts each other to understand their tasks, the consistent maintenance of integrated software and system requirement specifications should be definitely required. In this paper, for the purpose of systematically managing various sentences of requirement specifications, all of the sentences are partitioned into their clusters depending upon their syntactic similarities. Further, the proposed method automatically recommends a standard template close to an input sentence, which turns out to be syntactically incorrect, so that users are able to update the input sentence of requirement specifications. In the domain of TRain Open Software ARchitecture (TROSAR) platform, the accuracies for the classification of requirement specifications and the recommendation of standard template were measured with totally 500 template sentences of requirement specifications. In the experiment, the average accuracy of cluster classification for the set of sentences was higher than 99%, and the average accuracy of recommendation for the similar templates was higher than 93%.

This project related to develop and implementation of autonomous vision based mobile robot following human. Human tracking algorithm will be developed to allow a mobile robot to follow a human. A wireless camera will be used for image capturing, and Matlab will be use to process the image captured, followed by controlling the mobile robot to follow the human. This system will allow the robot to differentiate a human in a picture. The foreground and background will be separated and the foreground is used to determine the object whether it’s human or not. Then classification algorithm is applied to find the centroid of the human. This centroid is then compared with the center of the image to get the location of the human with respect to the camera, either at the left or right of the camera. If the human is not in the center of the camera view, then corrective measures is taken so that the human will be in the center of the camera view. Data for the centroid of human is shown through the Graphical User Interface (GUI).

Shape of granule is one of the important issues in granular computing classification problems and related to the classification accuracy, the number of granule, and the join process of two granules. A bottle up granular computing classification algorithm (BUGrC) is developed in the frame work of fuzzy lattices. Firstly, the granules are represented as 4 shapes, namely hyperdiamond granule, hypersphere granule, hypercube granule, and hyperbox granule. Secondly, the granule set is induced by the training set and the bottle up join operator. Thirdly, machine learning benchmark datasets are used to analyze and discuss the BUGrC with different shape granules.

A grouping of data objects such that the objects within a group are similar (or related) to one another and different from (or unrelated to) the objects in other groups. Many of clustering algorithm is available to analyze data. This paper intends to study and compare different clustering algorithms. These algorithms include K-Means, Farthest First, DBSCAN, CURE, Chameleon algorithm. All these algorithms are compared on the basis of their pros and cons, similarity measure, their working, functionality and time complexity.

In recent years, machine learning is increasingly becoming an important field of computer science. A new method using KNN classification algorithm identifies the layered boundary to find subgoal condition, to automatic classifying of large state space, reaches the dimension reduction of state space, and on the basis of generated subspace classifying to structure subtasks, and then realizes the hierarchical learning tasks automatically. In autonomous system, Agent assigns to their task through interaction with the environment, using hierarchical reinforcement learning technology can help the Agent in the large, complex environment to improve learning efficiency. Through the experimental results the effectiveness of the proposed algorithm is demonstrated. The goal of this paper is to provide a basic overview for both specialists and non-specialists to how to decide a good reinforcement learning algorithm for classification.

In recent years，machine learning method has been applied to the extensive research on traffic classification. In these methods, SVM (Support vector machine) is a supervised learning which can improve generalization ability of learning machine effectively. However, the penalty parameter C and kernel function parameter  are generally given by test experience during training of SVM. How to determine the optimal parameters of SVM is a problem to be solved. We proposed a method to deriving the optimal parameters of SVM based on GA (Genetic algorithm).This method does not need to traverse all the parameter points. The method extracts a certain number population from random solutions, and ultimately produces SVM optimal parameters according to the specific rules of operation. Through the method, we derived the optimal parameters combination C and  of SVM. The accuracy of network traffic classification is improved greatly.

Predicting crime types by using classification algorithms can help to find factors affecting crimes and prevent crimes. Due to various reasons in the process of data collection, there are often a large number of missing values in actual criminal dataset, which seriously affects the classification accuracy. Therefore, based on mutual KNNI (K nearest neighbor imputation) algorithm and combined with GRA (Grey Relational Analysis) theory, a novel data filling algorithm called GMKNN is proposed in order to improve the classification accuracy. The algorithm replaces the Euclidean distance formula used in mutual KNNI algorithm with the Grey relational grade formula to eliminate the effect of noise from the nearest neighbors and effectively deal with the discrete attributes. By comparing with several popular data filling algorithms based on a real criminal dataset with lots of missing values, higher classification accuracy can be obtained by using GMKNN algorithm, which is up to 77.837%.

This paper provides a comprehensive overview of UAV classification, tracking, and detection, offering researchers a clear understanding of these fundamental concepts. It elucidates how classification categorizes UAVs based on attributes, how tracking monitors real-time positions, and how detection identifies UAV presence. The interconnectedness of these aspects is highlighted, with detection enhancing tracking and classification aiding in anomaly identification. Moreover, the paper emphasizes the relevance of simulations in the context of drones and UAVs, underscoring their pivotal role in training, testing, and research. By succinctly presenting these core concepts and their practical implications, the paper equips researchers with a solid foundation to comprehend and explore the complexities of UAV operations and the role of simulations in advancing this dynamic field.

In this paper, the execution of some features based on wavelet transform are evaluated through classification of audio to speech and music using the MLP classifiers Optimized by Genetic Algorithm. Classification results show the wavelet features are completely successful in speech/music classification. Experimental comparisons using different wavelets are presented and discussed. By using some wavelet features, extracted from 1-second segments of the signal, we obtained 96.49% accuracy in the audio classification of the MLP classifiers optimized by genetic algorithm.

There are several statistical classification algorithms available for land use/land cover classification. However, each has a certain bias or compromise. Some methods like the parallel piped approach in supervised classification, cannot classify continuous regions within a feature. On the other hand, while unsupervised classification method takes maximum advantage of spectral variability in an image, the maximally separable clusters in spectral space may not do much for our perception of important classes in a given study area. In this research, the output of an ANN algorithm was compared with the Possibilistic c-Means an improvement of the fuzzy c-Means on both moderate resolutions Landsat8 and a high resolution Formosat 2 images. The Formosat 2 image comes with an 8m spectral resolution on the multispectral data. This multispectral image data was resampled to 10m in order to maintain a uniform ratio of 1:3 against Landsat 8 image. Six classes were chosen for analysis including: Dense forest, eucalyptus, water, grassland, wheat and riverine sand. Using a standard false color composite (FCC), the six features reflected differently in the infrared region with wheat producing the brightest pixel values. Signature collection per class was therefore easily obtained for all classifications. The output of both ANN and FCM, were analyzed separately for accuracy and an error matrix generated to assess the quality and accuracy of the classification algorithms. When you compare the results of the two methods on a per-class-basis, ANN had a crisper output compared to PCM which yielded clusters with pixels especially on the moderate resolution Landsat 8 imagery.

Classification in biomedical applications is an important task that predicts or classifies an outcome based on a given set of input variables such as diagnostic tests or the symptoms of a patient. Traditionally the classification algorithms would have to digest a stationary set of historical data in order to train up a decision-tree model and the learned model could then be used for testing new samples. However, a new breed of classification called stream-based classification can handle continuous data streams, which are ever evolving, unbound, and unstructured, for instance--biosignal live feeds. These emerging algorithms can potentially be used for real-time classification over biosignal data streams like EEG and ECG, etc. This paper presents a pioneer effort that studies the feasibility of classification algorithms for analyzing biosignals in the forms of infinite data streams. First, a performance comparison is made between traditional and stream-based classification. The results show that accuracy declines intermittently for traditional classification due to the requirement of model re-learning as new data arrives. Second, we show by a simulation that biosignal data streams can be processed with a satisfactory level of performance in terms of accuracy, memory requirement, and speed, by using a collection of stream-mining algorithms called Optimized Very Fast Decision Trees. The algorithms can effectively serve as a corner-stone technology for real-time classification in future biomedical applications.

To improve system performance, we apply the concept of membership function to Variance Considered Machines (VCMs) which is a modified algorithm of Support Vector Machines (SVMs) proposed in our previous studies. Many classification algorithms separate nonlinear data well. However, existing algorithms have ignored the fact that probabilities of error are very high in the data-mixed area. Therefore, we make our algorithm ignore data which has high error probabilities and consider data importantly which has low error probabilities to generate system output according to the probabilities of error. To get membership function, we calculate sigmoid function from the dataset by considering means and variances. After computation, this membership function is applied to the VCMs.

Database classification is an important preprocessing step for the multi-database mining (MDM). In fact, when a multi-branch company needs to explore its distributed data for decision making, it is imperative to classify these multiple databases into similar clusters before analyzing the data. To search for the best classification of a set of n databases, existing algorithms generate from 1 to ($n^2-n$)/2 candidate classifications. Although each candidate classification is included in the next one (i.e., clusters in the current classification are subsets of clusters in the next classification), existing algorithms generate each classification independently, that is, without taking into account the use of clusters from the previous classification. Consequently, existing algorithms are time consuming, especially when the number of candidate classifications increases. To overcome the latter problem, we propose in this paper an efficient approach that represents the problem of classifying the multiple databases as a problem of identifying the connected components of an undirected weighted graph. Theoretical analysis and experiments on public databases confirm the efficiency of our algorithm against existing works and that it overcomes the problem of increase in the execution time.