• The KIPS Transactions:PartB
• /
• v.16B no.3
• /
• pp.233-238
• /
• 2009

This paper presents an approach to the classifications of small size document using the instance-based feature filtering Relief-F algorithm. In the document classifications, we have not always good classification performances of small size document included a few features. Because total number of feature in the document set is large, but feature count of each document is very small relatively, so the similarities between documents are very low when we use general assessment of similarity and classifiers. Specially, in the cases of the classification of web document in the directory service and the classification of the sectors that cannot connect with the original file after recovery hard-disk, we have not good classification performances. Thus, we propose the Extended Relief-F(ERelief-F) algorithm using instance-based feature filtering algorithm Relief-F to solve problems of Relief-F as preprocess of classification. For the performance comparison, we tested information gain, odds ratio and Relief-F for feature filtering and getting those feature values, and used kNN and SVM classifiers. In the experimental results, the Extended Relief-F(ERelief-F) algorithm, compared with the others, performed best for all of the datasets and reduced many irrelevant features from document sets.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.8 no.5
• /
• pp.384-393
• /
• 2015

This paper proposes an optimized classifier based on feature elimination (OCFE) for gene selection with combining two feature elimination methods, ReliefF and SVM-RFE. ReliefF algorithm is filter feature selection which rank the data by the importance of the data. SVM-RFE algorithm is a wrapper feature selection which wrapped the data and rank the data based on the weight of feature. With combining these two methods we get less error rate average, 0.3016138 for OCFE and 0.3096779 for SVM-RFE. The proposed method also get better accuracy with 70% for OCFE and 69% for SVM-RFE.

• The Korean Journal of Applied Statistics
• /
• v.34 no.2
• /
• pp.153-166
• /
• 2021

A main goal of pharmacogenomics studies is to predict individual's drug responsiveness based on high dimensional genetic variables. Due to a large number of variables, feature selection is required in order to reduce the number of variables. The selected features are used to construct a predictive model using machine learning algorithms. In the present study, we applied several hybrid feature selection methods such as combinations of logistic regression, ReliefF, TurF, random forest, and LASSO to a next generation sequencing data set of 400 epilepsy patients. We then applied the selected features to machine learning methods including random forest, gradient boosting, and support vector machine as well as a stacking ensemble method. Our results showed that the stacking model with a hybrid feature selection of random forest and ReliefF performs better than with other combinations of approaches. Based on a 5-fold cross validation partition, the mean test accuracy value of the best model was 0.727 and the mean test AUC value of the best model was 0.761. It also appeared that the stacking models outperform than single machine learning predictive models when using the same selected features.

• Journal of Digital Convergence
• /
• v.17 no.3
• /
• pp.107-117
• /
• 2019

In the era of SNS, many people relies on it to express their emotions about various kinds of products and services. Therefore, for the companies eagerly seeking to investigate how their products and services are perceived in the market, emotion mining tasks using dataset from SNSs become important much more than ever. Basically, emotion mining is a branch of sentiment analysis which is based on BOW (bag-of-words) and TF-IDF. However, there are few studies on the emotion mining which adopt feature selection (FS) methods to look for optimal set of features ensuring better results. In this sense, this study aims to propose FS methods to conduct emotion mining tasks more effectively with better outcomes. This study uses Twitter and SemEval2007 dataset for the sake of emotion mining experiments. We applied three FS methods such as CFS (Correlation based FS), IG (Information Gain), and ReliefF. Emotion mining results were obtained from applying the selected features to nine classifiers. When applying DT (decision tree) to Tweet dataset, accuracy increases with CFS, IG, and ReliefF methods. When applying LR (logistic regression) to SemEval2007 dataset, accuracy increases with ReliefF method.

• Korean Journal of Optics and Photonics
• /
• v.9 no.2
• /
• pp.51-58
• /
• 1998

We investigated surface-relief profiles of the TPHK(telecentric paraxial holographic kinofrm) used as a Fourier-transform lens employing exact geometrical raytracing. For the TPHK of F/8 and focal length of 15 mm, we consider the cases where the thickness of the substrate is 0 and 50 ${\mu}{\textrm}{m}$, dividing the surface-relif profiles into fifty steps from plano-convex to convexplano shapes and varying the angle of incidence($0^{\circ},{2.5}^{\circ},5^{\circ}$). In order to identify appropriate surface-relief profiles, we employ, as criteria of performance, rms spot size, rms deviation from $f{\sin}{\theta}$, peak position and FWHM(full width at half maximum), number of rays abandoned from raytracing etc., which are determined from the result of exact raytracing. It is found that the profile with 80% of its relief thickness facing the image plane gives the best performance regardless of the presence of substrate.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2000.11a
• /
• pp.892-897
• /
• 2000

In this study, tap wear was analyzed not only to predict the tap life time but also to propose an improved tap design. Because rake angle and thread relief of tap are the significant factors in geometry of tap, these two factors were picked as the experiment variables. The experiment was conducted with six types of tap which have 6$^{\circ}$ , 8$^{\circ}$ and 12$^{\circ}$ of rake angles with 0,08mm and 0.14 mm of thread relief. From the measured cutting force, it could be known that cutting torque was low at the large the rake thread relief and tool life was long as rake angle became large. Eventually, tool life is longest when rake angle is 12$^{\circ}$ and the with 0.08mm thread relief. Aand the width of crater wear and that of flank wear were measured when a tapping torque was 20 [$kg_f$-mm] . Most of the measured values were above the width of tool wear[$V_B$=O.O3m], which means that tool life is over.

• Journal of the Korean Society of Safety
• /
• v.14 no.4
• /
• pp.93-99
• /
• 1999

A try was carried out to reduce lead time of die manufacturing. That is to make manual machining and finishing work time shorter by improving CAD/CAM modeling methods, so called with "separate modeling" and "relief modeling". The manual machining and finishing manual work time were reduced adapting the novel CAD/CAM modeling methods. Ultimately we accomplished much reduction of the lead time of die manufacturing.f die manufacturing.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.879-894
• /
• 2003

We evaluated the efficiencies of applying attribute selection methods and prior discretization to supervised learning, modelled by C4.5 and Naive Bayes. Three databases were obtained from UCI data archive, which consisted of continuous attributes except for one decision attribute. Four methods were used for attribute selection : MDI, ReliefF, Gain Ratio and Consistency-based method. MDI and ReliefF can be used for both continuous and discrete attributes, but the other two methods can be used only for discrete attributes. Discretization was performed using the Fayyad and Irani method. To investigate the effect of noise included in the database, noises were introduced into the data sets up to the extents of 10 or 20%, and then the data, including those either containing the noises or not, were processed through the steps of attribute selection, discretization and classification. The results of this study indicate that classification of the data based on selected attributes yields higher accuracy than in the case of classifying the full data set, and prior discretization does not lower the accuracy.

• Journal of Korean Academy of Nursing
• /
• v.36 no.6
• /
• pp.992-1001
• /
• 2006

Purpose: This study was done to provide data for a nursing intervention to alleviate newborn pain clinically by investigating the effect of oral glucose. Methods: Subjects were newborns hospitalized in the nursery. Informed consent was obtained from parents of 60 newborns. A heel stick was carried out for a test on 3 groups; the experimental, placebo, and control group. The Neonatal infant pain scale(NIPS), respiration rate, heart rate, peripheral oxygen partial pressure($SpO_2$), and crying duration were measured to assess pain reaction. All neonatal behaviors were recorded on videotape. Results: There were significant differences in pain behavior during stimulus(F=4.195, p=.020), pain behavior immediately after blood-sampling (F=4.114, p=.021), and pain behavior 3 minutes after that (F=3.630, p=.033). However, there were no significant differences in heart rate, respiration rate, peripheral oxygen partial pressure or crying duration after the heel stick among the groups. Conclusions: Oral administration of glucose before a heel stick caused the reduction of neonatal pain behavior, which means that it has an effect of pain relief.

• The Korean Journal of Pain
• /
• v.30 no.3
• /
• pp.214-219
• /
• 2017

Background: As herpes zoster progresses via postherpetic neuralgia (PHN) to well-established PHN, it presents its recalcitrant nature to the treatment. At this point, the well-established PHN is fixed as a non-treatable, but manageable chronic painful neuropathic disorder. This study evaluated the incidence of complete relief from PHN according to PHN duration at their first visit, and the other factors influencing it. Methods: A retrospective chart review was performed on patients with PHN at a university-based pain clinic over 7 years. The responders were defined as patients who had complete relief from pain after 1 year of active treatment. Age, sex, PHN duration at their first visit, dermatomal distribution, and underlying disorders were compared in the responder and non-responder groups. Responders were also compared according to these factors. Results: Among 117 PHN patients (M/F = 48/69), 35 patients (29.9%) had complete relief from PHN. Mean ages were $64.3{\pm}10.6$ and $66.9{\pm}10.7$ years, numbers of male to female patients were 11/24 and 37/45, and mean durations of PHN at their first visit were $8.5{\pm}6.3$ and $15.3{\pm}10.7$ months in the responder and non- responder groups, respectively. In addition, PHN patients who visited the clinic before 9 months showed a better result. Dermatomal distribution and underlying disorders did not show significant differences. Conclusions: Almost 30% of PHN patients received complete relief. Those who sought treatment in a pain clinic before 9 months received a better result.