• The Korean Journal of Nuclear Medicine Technology
• /
• v.20 no.2
• /
• pp.62-68
• /
• 2016

Purpose In a cyclosporine experiment using a robotic liquid handing system has found a deviation of its standard curve and low reproducibility of patients's results. The difference of the test is that methanol is mixed with samples and the extractions are used for the test. Therefore, we assumed that the abnormal test results came from using methanol and conducted this test. In a manual of a robotic liquid handling system mentions that we can choose several setting parameters depending on the viscosity of the liquids being used, the size of the sampling tips and the motor speeds that you elect to use but there's no exact order. This study was undertaken to confirm pipetting ability depending on types of liquids and investigate proper setting parameters for the optimum dispensing ability. Materials and Methods 4types of liquids(water, serum, methanol, PEG 6000(25%)) and $TSH^{125}I$ tracer(515 kBq) are used to confirm pipetting ability. 29 specimens for Cyclosporine test are used to compare results. Prepare 8 plastic tubes for each of the liquids and with multi pipette $400{\mu}l$ of each liquid is dispensed to 8 tubes and $100{\mu}l$ of $TSH^{125}I$ tracer are dispensed to all of the tubes. From the prepared samples, $100{\mu}l$ of liquids are dispensed using a robotic liquid handing system, counted and calculated its CV(%) depending on types of liquids. And then by adjusting several setting parameters(air gap, dispense time, delay time) the change of the CV(%)are calcutated and finds optimum setting parameters. 29 specimens are tested with 3 methods. The first(A) is manual method and the second(B) is used robotic liquid handling system with existing parameters. The third(C) is used robotic liquid handling system with adjusted parameters. Pipetting ability depending on types of liquids is assessed with CV(%). On the basis of (A), patients's test results are compared (A)and(B), (A)and(C) and they are assessed with %RE(%Relative error) and %Diff(%Difference). Results The CV(%) of the CPM depending on liquid types were water 0.88, serum 0.95, methanol 10.22 and PEG 0.68. As expected dispensing of methanol using a liquid handling system was the problem and others were good. The methanol's dispensing were conducted by adjusting several setting parameters. When transport air gap 0 was adjusted to 2 and 5, CV(%) were 20.16, 12.54 and when system air gap 0 was adjusted to 2 and 5, CV(%) were 8.94, 1.36. When adjusted to system air gap 2, transport air gap 2 was 12.96 and adjusted to system air gap 5, Transport air gap 5 was 1.33. When dispense speed was adjusted 300 to 100, CV(%) was 13.32 and when dispense delay was adjusted 200 to 100 was 13.55. When compared (B) to (A), the result increased 99.44% and %RE was 93.59%. When compared (C-system air gap was adjusted 0 to 5) to (A), the result increased 6.75% and %RE was 5.10%. Conclusion Adjusting speed and delay time of aspiration and dispense was meaningless but changing system air gap was effective. By adjusting several parameters proper value was found and it affected the practical result of the experiment. To optimize the system active efforts are needed through the test and in case of dispensing new types of liquids proper test is required to check the liquid is suitable for using the equipment.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.219-239
• /
• 2019

As the importance of providing customized services to individuals becomes important, researches on personalized recommendation systems are constantly being carried out. Collaborative filtering is one of the most popular systems in academia and industry. However, there exists limitation in a sense that recommendations were mostly based on quantitative information such as users' ratings, which made the accuracy be lowered. To solve these problems, many studies have been actively attempted to improve the performance of the recommendation system by using other information besides the quantitative information. Good examples are the usages of the sentiment analysis on customer review text data. Nevertheless, the existing research has not directly combined the results of the sentiment analysis and quantitative rating scores in the recommendation system. Therefore, this study aims to reflect the sentiments shown in the reviews into the rating scores. In other words, we propose a new algorithm that can directly convert the user 's own review into the empirically quantitative information and reflect it directly to the recommendation system. To do this, we needed to quantify users' reviews, which were originally qualitative information. In this study, sentiment score was calculated through sentiment analysis technique of text mining. The data was targeted for movie review. Based on the data, a domain specific sentiment dictionary is constructed for the movie reviews. Regression analysis was used as a method to construct sentiment dictionary. Each positive / negative dictionary was constructed using Lasso regression, Ridge regression, and ElasticNet methods. Based on this constructed sentiment dictionary, the accuracy was verified through confusion matrix. The accuracy of the Lasso based dictionary was 70%, the accuracy of the Ridge based dictionary was 79%, and that of the ElasticNet (${\alpha}=0.3$) was 83%. Therefore, in this study, the sentiment score of the review is calculated based on the dictionary of the ElasticNet method. It was combined with a rating to create a new rating. In this paper, we show that the collaborative filtering that reflects sentiment scores of user review is superior to the traditional method that only considers the existing rating. In order to show that the proposed algorithm is based on memory-based user collaboration filtering, item-based collaborative filtering and model based matrix factorization SVD, and SVD ++. Based on the above algorithm, the mean absolute error (MAE) and the root mean square error (RMSE) are calculated to evaluate the recommendation system with a score that combines sentiment scores with a system that only considers scores. When the evaluation index was MAE, it was improved by 0.059 for UBCF, 0.0862 for IBCF, 0.1012 for SVD and 0.188 for SVD ++. When the evaluation index is RMSE, UBCF is 0.0431, IBCF is 0.0882, SVD is 0.1103, and SVD ++ is 0.1756. As a result, it can be seen that the prediction performance of the evaluation point reflecting the sentiment score proposed in this paper is superior to that of the conventional evaluation method. In other words, in this paper, it is confirmed that the collaborative filtering that reflects the sentiment score of the user review shows superior accuracy as compared with the conventional type of collaborative filtering that only considers the quantitative score. We then attempted paired t-test validation to ensure that the proposed model was a better approach and concluded that the proposed model is better. In this study, to overcome limitations of previous researches that judge user's sentiment only by quantitative rating score, the review was numerically calculated and a user's opinion was more refined and considered into the recommendation system to improve the accuracy. The findings of this study have managerial implications to recommendation system developers who need to consider both quantitative information and qualitative information it is expect. The way of constructing the combined system in this paper might be directly used by the developers.

• Korean Journal of Agricultural Science
• /
• v.20 no.2
• /
• pp.167-180
• /
• 1993

Most of small size diesel engines are widely used with the same size and weight flywheel in the levels of 6.0kW and 7.5kW. This study was conducted to obtain basic data which affect the engine performance of the power tiller. The flywheel weight was considered as a major factor in this research. Basically, fuel consumption ratio, motoring loss, torque, vibration and mechanical efficiency of the engine were measured and analyzed on four levels of flywheel weight, 32.2, 29.4, 26.2 and $24.2kg_f$, respectively. Results were obtained as follows: 1. The weights of flywheel were $23.7kg_f$ from design program of JSME and $24.5kg_f$ from ASME and SAE design criteria. Therefore, the flywheel weight of $32.2kg_f$ might be reduced about $8kg_f$ in 7.5kW engine. 2. The rated outputs of 6.0kW and 7.5kW engine were actually 7.43kW and 7.85kW, respectively. When flywheel weight was reduced from $32.2kg_f$ to $24.2kg_f$, outputs were increased from 7.43kW to 7.70kW in 6.0kW engine and from 7.85kW to 8.25kW in 7.5kW engine. 3. When the flywheel weight was reduced from $32.2kg_f$ to $24.2kg_f$, fuel consumption ratio was decreased from 300.8 to 296.8g/kW-hr in 6.0kW engine and also from 313.6 to 312.8g/kW-hr in 7.5 kW engine, respectively. 4. When the flywheel weight was reduced from $32.2kg_f$ to $24.2kg_f$, mechanical efficiency of engine was increased from 76.1% to 76.8% in 6.0kW engine and also from 76.7% to 77.0% in 7.5kW engine, respectively. 5. When the flywheel weight was reduced from $32.2kg_f$ to $24.2kg_f$, vibration was decreased at X-axis and Z-axis in 6.0kW engine, however, slightly increased at Y-axis in 6.0kW engine and at all axes in 7.5kW engine. 6. When the flywheel weight was reduced from $32.2kg_f$ to $24.4kg_f$ motoring loss was decreased from 2.33kW to 1.75kW in 6.0kW engine and also from 2.46kW to 1.84kW in 7.5kW engine.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.13 no.1
• /
• pp.51-56
• /
• 2009

Purpose: Recently, most hospitals are introducing the PACS system and use of the system continues to expand. But small-scaled PACS called MicroPACS has already been in use through open source programs. The aim of this study is to prove utility of operating a MicroPACS, as a substitute back-up device for conventional storage media like CDs and DVDs, in addition to the full-PACS already in use. This study contains the way of setting up a MicroPACS with open source programs and assessment of its storage capability, stability, compatibility and performance of operations such as "retrieve", "query". Materials and Methods: 1. To start with, we searched open source software to correspond with the following standards to establish MicroPACS, (1) It must be available in Windows Operating System. (2) It must be free ware. (3) It must be compatible with PET/CT scanner. (4) It must be easy to use. (5) It must not be limited of storage capacity. (6) It must have DICOM supporting. 2. (1) To evaluate availability of data storage, we compared the time spent to back up data in the open source software with the optical discs (CDs and DVD-RAMs), and we also compared the time needed to retrieve data with the system and with optical discs respectively. (2) To estimate work efficiency, we measured the time spent to find data in CDs, DVD-RAMs and MicroPACS. 7 technologists participated in this study. 3. In order to evaluate stability of the software, we examined whether there is a data loss during the system is maintained for a year. Comparison object; How many errors occurred in randomly selected data of 500 CDs. Result: 1. We chose the Conquest DICOM Server among 11 open source software used MySQL as a database management system. 2. (1) Comparison of back up and retrieval time (min) showed the result of the following: DVD-RAM (5.13,2.26)/Conquest DICOM Server (1.49,1.19) by GE DSTE (p<0.001), CD (6.12,3.61)/Conquest (0.82,2.23) by GE DLS (p<0.001), CD (5.88,3.25)/Conquest (1.05,2.06) by SIEMENS. (2) The wasted time (sec) to find some data is as follows: CD ($156{\pm}46$), DVD-RAM ($115{\pm}21$) and Conquest DICOM Server ($13{\pm}6$). 3. There was no data loss (0%) for a year and it was stored 12741 PET/CT studies in 1.81 TB memory. In case of CDs, On the other hand, 14 errors among 500 CDs (2.8%) is generated. Conclusions: We found that MicroPACS could be set up with the open source software and its performance was excellent. The system built with open source proved more efficient and more robust than back-up process using CDs or DVD-RAMs. We believe that the operation of the MicroPACS would be effective data storage device as long as its operators develop and systematize it.

• Cartoon and Animation Studies
• /
• s.41
• /
• pp.455-486
• /
• 2015

The 3D printing technology, which started from the patent registration in 1986, was a technology that did not attract attention other than from some companies, due to the lack of awareness at the time. However, today, as expiring patents are appearing after the passage of 20 years, the price of 3D printers have decreased to the level of allowing purchase by individuals and the technology is attracting attention from industries, in addition to the general public, such as by naturally accepting 3D and to share 3D data, based on the generalization of online information exchange and improvement of computer performance. The production capability of 3D printers, which is based on digital data enabling digital transmission and revision and supplementation or production manufacturing not requiring molding, may provide a groundbreaking change to the process of manufacturing, and may attain the same effect in the character merchandise sector. Using a 3D printer is becoming a necessity in various figure merchandise productions which are in the forefront of the kidult culture that is recently gaining attention, and when predicting the demand by the industrial sites related to such character merchandise and when considering the more inexpensive price due to the expiration of patents and sharing of technology, expanding opportunities and sectors of employment and cultivating manpower that are able to engage in further creative work seems as a must, by introducing education courses cultivating manpower that can utilize 3D printers at the education field. However, there are limits in the information that can be obtained when seeking to introduce 3D printers in school education. Because the press or information media only mentions general information, such as the growth of the industrial size or prosperous future value of 3D printers, the research level of the academic world also remains at the level of organizing contents in an introductory level, such as by analyzing data on industrial size, analyzing the applicable scope in the industry, or introducing the printing technology. Such lack of information gives rise to problems at the education site. There would be no choice but to incur temporal and opportunity expenses, since the technology would only be able to be used after going through trials and errors, by first introducing the technology without examining the actual information, such as through comparing the strengths and weaknesses. In particular, if an expensive equipment introduced does not suit the features of school education, the loss costs would be significant. This research targeted general users without a technology-related basis, instead of specialists. By comparing the strengths and weaknesses and analyzing the problems and matters requiring notice upon use, pursuant to the representative technologies, instead of merely introducing the 3D printer technology as had been done previously, this research sought to explain the types of features that a 3D printer should have, in particular, when required in education relating to the development of figure merchandise as an optional cultural contents at cartoon-related departments, and sought to provide information that can be of practical help when seeking to provide education using 3D printers in the future. In the main body, the technologies were explained by making a classification based on a new perspective, such as the buttress method, types of materials, two-dimensional printing method, and three-dimensional printing method. The reason for selecting such different classification method was to easily allow mutual comparison of the practical problems upon use. In conclusion, the most suitable 3D printer was selected as the printer in the FDM method, which is comparatively cheap and requires low repair and maintenance cost and low materials expenses, although rather insufficient in the quality of outputs, and a recommendation was made, in addition, to select an entity that is supportive in providing technical support.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.79-104
• /
• 2020

Recently, as deep learning has attracted attention, the use of deep learning is being considered as a method for solving problems in various fields. In particular, deep learning is known to have excellent performance when applied to applying unstructured data such as text, sound and images, and many studies have proven its effectiveness. Owing to the remarkable development of text and image deep learning technology, interests in image captioning technology and its application is rapidly increasing. Image captioning is a technique that automatically generates relevant captions for a given image by handling both image comprehension and text generation simultaneously. In spite of the high entry barrier of image captioning that analysts should be able to process both image and text data, image captioning has established itself as one of the key fields in the A.I. research owing to its various applicability. In addition, many researches have been conducted to improve the performance of image captioning in various aspects. Recent researches attempt to create advanced captions that can not only describe an image accurately, but also convey the information contained in the image more sophisticatedly. Despite many recent efforts to improve the performance of image captioning, it is difficult to find any researches to interpret images from the perspective of domain experts in each field not from the perspective of the general public. Even for the same image, the part of interests may differ according to the professional field of the person who has encountered the image. Moreover, the way of interpreting and expressing the image also differs according to the level of expertise. The public tends to recognize the image from a holistic and general perspective, that is, from the perspective of identifying the image's constituent objects and their relationships. On the contrary, the domain experts tend to recognize the image by focusing on some specific elements necessary to interpret the given image based on their expertise. It implies that meaningful parts of an image are mutually different depending on viewers' perspective even for the same image. So, image captioning needs to implement this phenomenon. Therefore, in this study, we propose a method to generate captions specialized in each domain for the image by utilizing the expertise of experts in the corresponding domain. Specifically, after performing pre-training on a large amount of general data, the expertise in the field is transplanted through transfer-learning with a small amount of expertise data. However, simple adaption of transfer learning using expertise data may invoke another type of problems. Simultaneous learning with captions of various characteristics may invoke so-called 'inter-observation interference' problem, which make it difficult to perform pure learning of each characteristic point of view. For learning with vast amount of data, most of this interference is self-purified and has little impact on learning results. On the contrary, in the case of fine-tuning where learning is performed on a small amount of data, the impact of such interference on learning can be relatively large. To solve this problem, therefore, we propose a novel 'Character-Independent Transfer-learning' that performs transfer learning independently for each character. In order to confirm the feasibility of the proposed methodology, we performed experiments utilizing the results of pre-training on MSCOCO dataset which is comprised of 120,000 images and about 600,000 general captions. Additionally, according to the advice of an art therapist, about 300 pairs of 'image / expertise captions' were created, and the data was used for the experiments of expertise transplantation. As a result of the experiment, it was confirmed that the caption generated according to the proposed methodology generates captions from the perspective of implanted expertise whereas the caption generated through learning on general data contains a number of contents irrelevant to expertise interpretation. In this paper, we propose a novel approach of specialized image interpretation. To achieve this goal, we present a method to use transfer learning and generate captions specialized in the specific domain. In the future, by applying the proposed methodology to expertise transplant in various fields, we expected that many researches will be actively conducted to solve the problem of lack of expertise data and to improve performance of image captioning.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.127-148
• /
• 2020

The data center is a physical environment facility for accommodating computer systems and related components, and is an essential foundation technology for next-generation core industries such as big data, smart factories, wearables, and smart homes. In particular, with the growth of cloud computing, the proportional expansion of the data center infrastructure is inevitable. Monitoring the health of these data center facilities is a way to maintain and manage the system and prevent failure. If a failure occurs in some elements of the facility, it may affect not only the relevant equipment but also other connected equipment, and may cause enormous damage. In particular, IT facilities are irregular due to interdependence and it is difficult to know the cause. In the previous study predicting failure in data center, failure was predicted by looking at a single server as a single state without assuming that the devices were mixed. Therefore, in this study, data center failures were classified into failures occurring inside the server (Outage A) and failures occurring outside the server (Outage B), and focused on analyzing complex failures occurring within the server. Server external failures include power, cooling, user errors, etc. Since such failures can be prevented in the early stages of data center facility construction, various solutions are being developed. On the other hand, the cause of the failure occurring in the server is difficult to determine, and adequate prevention has not yet been achieved. In particular, this is the reason why server failures do not occur singularly, cause other server failures, or receive something that causes failures from other servers. In other words, while the existing studies assumed that it was a single server that did not affect the servers and analyzed the failure, in this study, the failure occurred on the assumption that it had an effect between servers. In order to define the complex failure situation in the data center, failure history data for each equipment existing in the data center was used. There are four major failures considered in this study: Network Node Down, Server Down, Windows Activation Services Down, and Database Management System Service Down. The failures that occur for each device are sorted in chronological order, and when a failure occurs in a specific equipment, if a failure occurs in a specific equipment within 5 minutes from the time of occurrence, it is defined that the failure occurs simultaneously. After configuring the sequence for the devices that have failed at the same time, 5 devices that frequently occur simultaneously within the configured sequence were selected, and the case where the selected devices failed at the same time was confirmed through visualization. Since the server resource information collected for failure analysis is in units of time series and has flow, we used Long Short-term Memory (LSTM), a deep learning algorithm that can predict the next state through the previous state. In addition, unlike a single server, the Hierarchical Attention Network deep learning model structure was used in consideration of the fact that the level of multiple failures for each server is different. This algorithm is a method of increasing the prediction accuracy by giving weight to the server as the impact on the failure increases. The study began with defining the type of failure and selecting the analysis target. In the first experiment, the same collected data was assumed as a single server state and a multiple server state, and compared and analyzed. The second experiment improved the prediction accuracy in the case of a complex server by optimizing each server threshold. In the first experiment, which assumed each of a single server and multiple servers, in the case of a single server, it was predicted that three of the five servers did not have a failure even though the actual failure occurred. However, assuming multiple servers, all five servers were predicted to have failed. As a result of the experiment, the hypothesis that there is an effect between servers is proven. As a result of this study, it was confirmed that the prediction performance was superior when the multiple servers were assumed than when the single server was assumed. In particular, applying the Hierarchical Attention Network algorithm, assuming that the effects of each server will be different, played a role in improving the analysis effect. In addition, by applying a different threshold for each server, the prediction accuracy could be improved. This study showed that failures that are difficult to determine the cause can be predicted through historical data, and a model that can predict failures occurring in servers in data centers is presented. It is expected that the occurrence of disability can be prevented in advance using the results of this study.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.15 no.2
• /
• pp.116-124
• /
• 2011

Purpose: To establish accurate external quality assurance (EQA) test, cross institutional and modality tests were performed using WHO certificated reference material (CRM) and same pooled patients serum. Materials and Methods: Accuracy and precision were evaluated using CRM and pooled patients' serum for AFP, CEA, PSA, CA 125, CA 19-9, T3, T4, Tg, TSH. To evaluate the accuracy and precision, recover test and coefficient variation were measured. RIA test were performed in major 5 RIA laboratory and EIA (CLIA) test were done in 5 major EIA laboratory. same sample of CRM and pooled serum were delivered to each laboratory. Results: In 2009, mean precision of total tumor marker of RIA was $14.8{\pm}4.2%$ and that of EIA(CLIA) was $19.2{\pm}6.9%$. In 2010, mean precision of 5 tumor marker and T3, T4, Tg, TSH was $13.8{\pm}6.1%$ in RIA and $15.5{\pm}7.7%$ in EIA (CLIA). There was no significant difference between RIA and EIA. In RIA, the coefficient variations (CV) of AFP, CEA, PSA, CA 125, T3, T4, TSH were within 20%. The CV of CA 19-9 was over 20% but there was no significant difference with EIA (CLIA) (p=0.345). In recovery test using CRM, AFP, PSA, T4, TSH showed 92~103% of recovery in RIA. In recovery test using commercial material, CEA, CA 125, CA 19-9 showed relatively lower recovery than CRM but there was no significant difference between RIA and EIA (CLIA). Conclusion: By evaluating the precision and accuracy of each test, EQA test could more accurately measured the quality of each test and performance of laboratory.

• Journal of Intelligence and Information Systems
• /
• v.22 no.1
• /
• pp.139-157
• /
• 2016

Bankruptcy involves considerable costs, so it can have significant effects on a country's economy. Thus, bankruptcy prediction is an important issue. Over the past several decades, many researchers have addressed topics associated with bankruptcy prediction. Early research on bankruptcy prediction employed conventional statistical methods such as univariate analysis, discriminant analysis, multiple regression, and logistic regression. Later on, many studies began utilizing artificial intelligence techniques such as inductive learning, neural networks, and case-based reasoning. Currently, ensemble models are being utilized to enhance the accuracy of bankruptcy prediction. Ensemble classification involves combining multiple classifiers to obtain more accurate predictions than those obtained using individual models. Ensemble learning techniques are known to be very useful for improving the generalization ability of the classifier. Base classifiers in the ensemble must be as accurate and diverse as possible in order to enhance the generalization ability of an ensemble model. Commonly used methods for constructing ensemble classifiers include bagging, boosting, and random subspace. The random subspace method selects a random feature subset for each classifier from the original feature space to diversify the base classifiers of an ensemble. Each ensemble member is trained by a randomly chosen feature subspace from the original feature set, and predictions from each ensemble member are combined by an aggregation method. The k-nearest neighbors (KNN) classifier is robust with respect to variations in the dataset but is very sensitive to changes in the feature space. For this reason, KNN is a good classifier for the random subspace method. The KNN random subspace ensemble model has been shown to be very effective for improving an individual KNN model. The k parameter of KNN base classifiers and selected feature subsets for base classifiers play an important role in determining the performance of the KNN ensemble model. However, few studies have focused on optimizing the k parameter and feature subsets of base classifiers in the ensemble. This study proposed a new ensemble method that improves upon the performance KNN ensemble model by optimizing both k parameters and feature subsets of base classifiers. A genetic algorithm was used to optimize the KNN ensemble model and improve the prediction accuracy of the ensemble model. The proposed model was applied to a bankruptcy prediction problem by using a real dataset from Korean companies. The research data included 1800 externally non-audited firms that filed for bankruptcy (900 cases) or non-bankruptcy (900 cases). Initially, the dataset consisted of 134 financial ratios. Prior to the experiments, 75 financial ratios were selected based on an independent sample t-test of each financial ratio as an input variable and bankruptcy or non-bankruptcy as an output variable. Of these, 24 financial ratios were selected by using a logistic regression backward feature selection method. The complete dataset was separated into two parts: training and validation. The training dataset was further divided into two portions: one for the training model and the other to avoid overfitting. The prediction accuracy against this dataset was used to determine the fitness value in order to avoid overfitting. The validation dataset was used to evaluate the effectiveness of the final model. A 10-fold cross-validation was implemented to compare the performances of the proposed model and other models. To evaluate the effectiveness of the proposed model, the classification accuracy of the proposed model was compared with that of other models. The Q-statistic values and average classification accuracies of base classifiers were investigated. The experimental results showed that the proposed model outperformed other models, such as the single model and random subspace ensemble model.