• Proceedings of the Korean Statistical Society Conference
• /
• 2005.11a
• /
• pp.193-197
• /
• 2005

Classification tree is one of the most suitable base learners for ensemble. For past decade, it was found that bagging gives the most accurate prediction when used with unpruned tree and boosting with stump. Researchers have tried to understand the relationship between the size of trees and the accuracy of ensemble. With experiment, it is found that large trees make boosting overfit the dataset and stumps help avoid it. It means that the accuracy of each classifier needs to be sacrificed for better weighting at each iteration. Hence, split effect in boosting can be explained with the trade-off between the accuracy of each classifier and better weighting on the misclassified points. In bagging, combining larger trees give more accurate prediction because bagging does not have such trade-off, thus it is advisable to make each classifier as accurate as possible.

• Journal of Intelligence and Information Systems
• /
• v.10 no.2
• /
• pp.133-145
• /
• 2004

Combining multiple classifiers to obtain improved performance over the individual classifier has been a widely used technique. The task of constructing a multiple classifier system(MCS) contains two different issues : how to generate a diverse set of base-level classifiers and how to combine their predictions. In this paper, we review the characteristics of the existing multiple classifier systems: bagging, boosting, and stacking. And then we propose new MCSs: stacked bagging, stacked boosting, bagged stacking, and boasted stacking. These MCSs are a sort of hybrid MCSs that combine advantageous characteristics of the existing ones. In order to evaluate the performance of the proposed schemes, we conducted experiments with nine different real-world datasets from UCI KDD archive. The result of experiments showed the superiority of our hybrid MCSs, especially bagged stacking and boosted stacking, over the existing ones.

• IMMUNE NETWORK
• /
• v.3 no.2
• /
• pp.110-117
• /
• 2003

Background: The usefulness of DNA vaccine at priming step of heterologous prime-boost vaccination led to DNA vaccine closer to practical reality. DNA vaccine priming followed by recombinant viral vector boosting via systemic route induces optimal systemic immunity but no mucosal immunity. Mucosal vaccination of the reversed protocol (recombinant viral vector priming-DNA vaccine boosting), however, can induce both maximal mucosal and systemic immunity. Here, we tried to address the reason why the mucosal protocol of prime-boost vaccination differs from that of systemic vaccination. Methods: To address the importance of primary immunity induced at priming step, mice were primed with different doses of DNA vaccine or coadministration of DNA vaccine plus mucosal adjuvant, and immunity including serum IgG and mucosal IgA was then determined following boosting with recombinant viral vector. Next, to assess influence of humoral pre-existing immunity on boosting $CD8^+$ T cell-mediated immunity, $CD8^+$ T cell-mediated immunity in B cell-deficient (${\mu}K/O$) mice immunized with prime-boost regimens was evaluated by CTL assay and $IFN-{\gamma}$-producing cells. Results: Immunity primed with recombinant viral vector was effectively boosted with DNA vaccine even 60 days later. In particular, animals primed by increasing doses of DNA vaccine or incorporating an adjuvant at priming step and boosted by recombinant viral vector elicited comparable responses to recombinant viral vector primed-DNA vaccine boosted group. Humoral pre-existing immunity was also unlikely to interfere the boosting effect of $CD8^+$ T cell-mediated immunity by recombinant viral vector. Conclusion: This report provides the important point that optimally primed responses should be considered in mucosal immunization of heterologous prime-boost regimens for inducing the effective boosting at both mucosal and systemic sites.

• The KIPS Transactions:PartB
• /
• v.11B no.5
• /
• pp.545-554
• /
• 2004

Combining multiple classifiers to obtain improved performance over the individual classifier has been a widely used technique. The task of constructing a multiple classifier system(MCS) contains two different Issues how to generate a diverse set of base-level classifiers and how to combine their predictions. In this paper, we review the characteristics of existing multiple classifier systems : Bagging, Boosting, and Slaking. For document classification, we propose new MCSs such as Stacked Bagging, Stacked Boosting, Bagged Stacking, Boosted Stacking. These MCSs are a sort of hybrid MCSs that combine advantages of existing MCSs such as Bugging, Boosting, and Stacking. We conducted some experiments of document classification to evaluate the performances of the proposed schemes on MEDLINE, Usenet news, and Web document collections. The result of experiments demonstrate the superiority of our hybrid MCSs over the existing ones.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.11
• /
• pp.53-63
• /
• 2023

This study developed a cognitive impairment predictive model as one of the screening tests for preventing dementia in the elderly by using Automated Machine Learning(AutoML). We used 'Wearable lifelog data for high-risk dementia patients' of National Information Society Agency, then conducted using PyCaret 3.0.0 in the Google Colaboratory environment. This study analysis steps are as follows; first, selecting five models demonstrating excellent classification performance for the model development and lifelog data analysis. Next, using ensemble learning to integrate these models and assess their performance. It was found that Voting Classifier, Gradient Boosting Classifier, Extreme Gradient Boosting, Light Gradient Boosting Machine, Extra Trees Classifier, and Random Forest Classifier model showed high predictive performance in that order. This study findings, furthermore, emphasized on the the crucial importance of 'Average respiration per minute during sleep' and 'Average heart rate per minute during sleep' as the most critical feature variables for accurate predictions. Finally, these study results suggest that consideration of the possibility of using machine learning and lifelog as a means to more effectively manage and prevent cognitive impairment in the elderly.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.5
• /
• pp.949-959
• /
• 2012

This article is to find the right size of decision trees that performs better for boosting algorithm. First we defined the tree size D as the depth of a decision tree. Then we compared the performance of boosting algorithm with different tree sizes in the experiment. Although it is an usual practice to set the tree size in boosting algorithm to be small, we figured out that the choice of D has a significant influence on the performance of boosting algorithm. Furthermore, we found out that the tree size D need to be sufficiently large for some dataset. The experiment result shows that there exists an optimal D for each dataset and choosing the right size D is important in improving the performance of boosting. We also tried to find the model for estimating the right size D suitable for boosting algorithm, using variables that can explain the nature of a given dataset. The suggested model reveals that the optimal tree size D for a given dataset can be estimated by the error rate of stump tree, the number of classes, the depth of a single tree, and the gini impurity.

• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.570-578
• /
• 2003

Maximum entropy models are promising candidates for natural language modeling. However, there are two major hurdles in applying maximum entropy models to real-life language problems, such as prepositional phrase attachment: feature selection and high computational complexity. In this paper, we propose a maximum entropy boosting model to overcome these limitations and the problem of imbalanced data in natural language resources, and apply it to prepositional phrase (PP) attachment and part-of-speech (POS) tagging. According to the experimental results on Wall Street Journal corpus, the model shows 84.3% of accuracy for PP attachment and 96.78% of accuracy for POS tagging that are close to the state-of-the-art performance of these tasks only with small efforts of modeling.

• Proceedings of the KIPE Conference
• /
• 2008.06a
• /
• pp.265-267
• /
• 2008

As generally recognized, the driving performance of a SRM at higher speed will be degraded due to the effects of back electromagnetic force (EMF). This phenomenon can be improved via voltage boosting. So in this paper an improved converter of enhancing the performance for four-phase switched reluctance motor (SRM) is proposed. By using one additional capacitor and switches, an extra controllable boosted voltage can be produced during the rise and fall periods of a motor phase current. Then this active boosted voltage can reduce the effect of EMF on the current, particularly at high speeds. The attractive features of the proposed converter are as follows: obtaining boosted voltage to improve performance of SRM with same numbers of switch and diode as asymmetric converter, having higher control flexibility and capability of boosting voltage compared with passive boosting converters, possessing lower cost and simple control in comparison with existing active boosting converters. The performances of the proposed circuit are verified by the simulation and experiment results.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.11
• /
• pp.5436-5458
• /
• 2017

Automatic Video Surveillance System (AVSS) has become important to computer vision researchers as crime has increased in the twenty-first century. As a new branch of AVSS, baggage detection has a wide area of security applications. Some of them are, detecting baggage in baggage restricted super shop, detecting unclaimed baggage in public space etc. However, in this paper, a detection & classification framework of baggage is proposed. Initially, background subtraction is performed instead of sliding window approach to speed up the system and HSI model is used to deal with different illumination conditions. Then, a model is introduced to overcome shadow effect. Then, occlusion of objects is detected using proposed mirroring algorithm to track individual objects. Extraction of rotational signal descriptor (SP-RSD-HOG) with support plane from Region of Interest (ROI) add rotation invariance nature in HOG. Finally, dynamic human body parameter setting approach enables the system to detect & classify single or multiple pieces of carried baggage even if some portions of human are absent. In baggage detection, a strong classifier is generated by boosting similarity measure based multi layer Support Vector Machine (SVM)s into HOG based SVM. This boosting technique has been used to deal with various texture patterns of baggage. Experimental results have discovered the system satisfactorily accurate and faster comparative to other alternatives.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.22 no.1
• /
• pp.87-94
• /
• 2008

A method for improving current THD of Valley-fill rectifier is proposed in this paper. The proposed topology combines a boosting inductor with Valley-fill rectifier which carry out AC/DC conversion and PFC simultaneously The boosting effect by PWM switching makes low THD current and improve of Valley-fill rectifier. The operation modes and THD of input current are analyzed as applied the boosting inductor, and the optimum value of boosting inductor is determined A 100[W] single-stage converter has been designed and tested. Experimental results are resented to verify the validity of the proposed method.