• 한국산림과학회지
• /
• 제111권2호
• /
• pp.302-310
• /
• 2022

본 연구는 효율적인 산림시업계획 수립을 위하여 입목의 측정정보와 머신러닝 알고리즘을 이용하여 주요 침엽수종(소나무, 잣나무, 낙엽송)의 수관급 분류를 목적으로 하였다. 입목의 측정정보는 9년간 수집된 국유림 모니터링 정보를 활용하였으며, 머신러닝 알고리즘은 Random Forest (RF), XGBoost (XGB), Light GBM (LGBM)을 사용하였다. 알고리즘의 정확도, 정밀도, 재현율, F1 score를 이용한 성능평가를 통하여 알고리즘의 정확도를 비교·평가하였다. 분석결과, 소나무림, 잣나무림, 낙엽송 모두 RF 알고리즘이 성능평가 점수가 가장 높았으며, 수종별로는 소나무가 정확도 약 65%, 정밀도 약 72%, 재현율 약 60%, F1 score 약 66%로 성능평가 점수가 가장 높았다. 수관급은 우세목의 정확도가 약 80%이상으로 높았으나, 준우세목과 중간목, 피압목의 분류 정확도는 낮게 평가되었다. 본 연구결과는 산림시업의 간벌목 선정에 있어 의사결정을 위한 참조자료로 활용이 가능할 것으로 판단된다.

• 한국비블리아학회지
• /
• 제11권2호
• /
• pp.5-23
• /
• 2000

본 연구는 지식과 분류 사이의 연관성을 고찰하였다. 분류표는 분류 대상의 지식을 표현하는 구조로 그 안의 포함되는 내용과 다양한 관계를 표현해 주는데 대표적인 4가지 분류학적 접근방식인 계층 구조, 나무 구조, 패러다임. 패싯 분석을 새로운 지식의 창조와 발견의 측면에서 장단점을 기술하였다. 이를 바탕으로 지식과 분류 과정이 서로 영향을 주고받는 방식을 확인하고, 미래의 보다나은 새로운 분류 방식의 필요성과 그 가능성을 제시하였다.

• Industrial Engineering and Management Systems
• /
• 제4권1호
• /
• pp.102-108
• /
• 2005

This paper studies medical data classification methods, comparing decision tree and system reconstruction analysis as applied to heart disease medical data mining. The data we study is collected from patients with coronary heart disease. It has 1,723 records of 71 attributes each. We use the system-reconstruction method to weight it. We use decision tree algorithms, such as induction of decision trees (ID3), classification and regression tree (C4.5), classification and regression tree (CART), Chi-square automatic interaction detector (CHAID), and exhausted CHAID. We use the results to compare the correction rate, leaf number, and tree depth of different decision-tree algorithms. According to the experiments, we know that weighted data can improve the correction rate of coronary heart disease data but has little effect on the tree depth and leaf number.

• 대한원격탐사학회지
• /
• 제22권3호
• /
• pp.211-219
• /
• 2006

A random forests classifier is applied to multi-sensor data fusion for supervised land-cover classification in order to account for the importance of variable. The random forests approach is a non-parametric ensemble classifier based on CART-like trees. The distinguished feature is that the importance of variable can be estimated by randomly permuting the variable of interest in all the out-of-bag samples for each classifier. Two different multi-sensor data sets for supervised classification were used to illustrate the applicability of random forests: one with optical and polarimetric SAR data and the other with multi-temporal Radarsat-l and ENVISAT ASAR data sets. From the experimental results, the random forests approach could extract important variables or bands for land-cover discrimination and showed reasonably good performance in terms of classification accuracy.

• 한국산림과학회지
• /
• 제97권5호
• /
• pp.501-507
• /
• 2008

The data of crown width with 4 directions, DBH, tree height, and coordinate for sample trees were collected from 30 permanent sample plots in secondary fore st of the Maoershan Experimental Forestry Farm, Northeast China. In this paper, the competition of individual trees in stand were discussed for secondary forest by using iterative Hegyi competition index and crown overlap index that represented the competitive and cooperative interactions among neighboring trees. Active competitors of subject tree in the competition zone were selected to calculate the iterative competition index. Using the results of crown classification based on the equal crown projection area, a new distance dependent competition index called crown overlap index (COI) was developed for secondary forest. The COI performed well in describing the crown competition rather than crown competition factor (CCF). The individual-based competition index discussed in this paper will provide more precise for developing individual tree growth models for secondary forest and it can also use to adjust the stand structure for spatial optimal management.

• Journal of Information Processing Systems
• /
• 제7권4호
• /
• pp.717-732
• /
• 2011

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.

• Communications for Statistical Applications and Methods
• /
• 제10권3호
• /
• pp.845-857
• /
• 2003

We explore the use of genetic programming to evolve decision trees directly for classification problems with both discrete and continuous predictors. We demonstrate that the derived hypotheses of standard algorithms can substantially deviated from the optimum. This deviation is partly due to their top-down style procedures. The performance of the system is measured on a set of real and simulated data sets and compared with the performance of well-known algorithms like CHAID, CART, C5.0, and QUEST. Proposed algorithm seems to be effective in handling problems caused by top-down style procedures of existing algorithms.

• Communications for Statistical Applications and Methods
• /
• 제10권3호
• /
• pp.627-635
• /
• 2003

In this short communication we discuss the bias problem of C4.5 in split variable selection and suggest a method to reduce the variable selection bias among categorical predictor variables. A penalty proportional to the number of categories is applied to the splitting criterion gain of C4.5. The results of empirical comparisons show that the proposed modification of C4.5 reduces the size of classification trees.

• 산업경영시스템학회지
• /
• 제24권63호
• /
• pp.23-31
• /
• 2001

Data mining can be either directed or undirected. One way of thinking about it is that we use undirected data mining to recognize relationship in the data and directed data mining to explain those relationships once they have been found. Several data mining techniques have received considerable research attention. In this paper, we propose an algorithm for discovering association rules as directed data mining and applying them to classification. In the first phase, we find frequent closed itemsets and association rules. After this phase, we construct the decision trees using discovered association rules. The algorithm can be applicable to customer relationship management.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.23.4-23
• /
• 2001

This paper proposes a new approach to unsupervised classification of remotely sensed images. Fusion of optic images (Landsat TM) and radar data (SAR) has beer used to increase the accuracy of classification. Number of clusters is estimated using generalized Dunns measure. Performance of the proposed method is best observed comparing the classified images with classified aerial images.