• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.23-45
• /
• 2020

Big data is creating in a wide variety of fields such as medical care, manufacturing, logistics, sales site, SNS, and the dataset characteristics are also diverse. In order to secure the competitiveness of companies, it is necessary to improve decision-making capacity using a classification algorithm. However, most of them do not have sufficient knowledge on what kind of classification algorithm is appropriate for a specific problem area. In other words, determining which classification algorithm is appropriate depending on the characteristics of the dataset was has been a task that required expertise and effort. This is because the relationship between the characteristics of datasets (called meta-features) and the performance of classification algorithms has not been fully understood. Moreover, there has been little research on meta-features reflecting the characteristics of multi-class. Therefore, the purpose of this study is to empirically analyze whether meta-features of multi-class datasets have a significant effect on the performance of classification algorithms. In this study, meta-features of multi-class datasets were identified into two factors, (the data structure and the data complexity,) and seven representative meta-features were selected. Among those, we included the Herfindahl-Hirschman Index (HHI), originally a market concentration measurement index, in the meta-features to replace IR(Imbalanced Ratio). Also, we developed a new index called Reverse ReLU Silhouette Score into the meta-feature set. Among the UCI Machine Learning Repository data, six representative datasets (Balance Scale, PageBlocks, Car Evaluation, User Knowledge-Modeling, Wine Quality(red), Contraceptive Method Choice) were selected. The class of each dataset was classified by using the classification algorithms (KNN, Logistic Regression, Nave Bayes, Random Forest, and SVM) selected in the study. For each dataset, we applied 10-fold cross validation method. 10% to 100% oversampling method is applied for each fold and meta-features of the dataset is measured. The meta-features selected are HHI, Number of Classes, Number of Features, Entropy, Reverse ReLU Silhouette Score, Nonlinearity of Linear Classifier, Hub Score. F1-score was selected as the dependent variable. As a result, the results of this study showed that the six meta-features including Reverse ReLU Silhouette Score and HHI proposed in this study have a significant effect on the classification performance. (1) The meta-features HHI proposed in this study was significant in the classification performance. (2) The number of variables has a significant effect on the classification performance, unlike the number of classes, but it has a positive effect. (3) The number of classes has a negative effect on the performance of classification. (4) Entropy has a significant effect on the performance of classification. (5) The Reverse ReLU Silhouette Score also significantly affects the classification performance at a significant level of 0.01. (6) The nonlinearity of linear classifiers has a significant negative effect on classification performance. In addition, the results of the analysis by the classification algorithms were also consistent. In the regression analysis by classification algorithm, Naïve Bayes algorithm does not have a significant effect on the number of variables unlike other classification algorithms. This study has two theoretical contributions: (1) two new meta-features (HHI, Reverse ReLU Silhouette score) was proved to be significant. (2) The effects of data characteristics on the performance of classification were investigated using meta-features. The practical contribution points (1) can be utilized in the development of classification algorithm recommendation system according to the characteristics of datasets. (2) Many data scientists are often testing by adjusting the parameters of the algorithm to find the optimal algorithm for the situation because the characteristics of the data are different. In this process, excessive waste of resources occurs due to hardware, cost, time, and manpower. This study is expected to be useful for machine learning, data mining researchers, practitioners, and machine learning-based system developers. The composition of this study consists of introduction, related research, research model, experiment, conclusion and discussion.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.3D
• /
• pp.419-431
• /
• 2009

The core of this study is finding out the efficient band selection or extraction method discovering the optimal spectral bands when applying canonical correlation classifier (CCC) to hyperspectral data. The optimal efficient bands grounded on each separability decision technique are selected using Multispec$^{(C)}$ software developed by Purdue university of USA. Total 6 separability decision techniques are used, which are Divergence, Transformed Divergence, Bhattacharyya, Mean Bhattacharyya, Covariance Bhattacharyya, Noncovariance Bhattacharyya. For feature extraction, PCA transformation and MNF transformation are accomplished by ERDAS Imagine and ENVI software. For the comparison and assessment on the effect of feature selection and feature extraction, land cover classification is performed by CCC. The overall accuracy of CCC using the firstly selected 60 bands is 71.8%, the highest classification accuracy acquired by CCC is 79.0% as the case that executes CCC after appling Noncovariance Bhattacharyya. In conclusion, as a matter of fact, only Noncovariance Bhattacharyya separability decision method was valuable as feature selection algorithm for hyperspectral image classification depended on CCC. The lassification accuracy using other feature selection and extraction algorithms except Divergence rather declined in CCC.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.44 no.1
• /
• pp.26-32
• /
• 2007

In sequential speaker classification, the lack of prior information about the speakers poses a challenge for model initialization. To address the challenge, a predetermined generic model set, called Sample Speaker Models, was previously proposed. This approach can be useful for accurate speaker modeling without requiring initial speaker data. However, an optimal method for sampling the models from a generic model pool is still required. To solve this problem, the Speaker Quantization method, motivated by vector quantization, is proposed. Experimental results showed that the new approach outperformed the random sampling approach with 25% relative improvement in error rate on switchboard telephone conversations.

• Proceedings of the KIEE Conference
• /
• 1997.11a
• /
• pp.600-602
• /
• 1997

In this study, a hybrid neural net consisting of an Adaptive LVQ(ALVQ) algorithm and MLP is proposed to perform speaker identification task. ALVQ is a new learning procedure using adaptively feature vector sequence instead of only one feature vector in training codebooks initialized by LBG algorithm and the optimization criterion of this method is consistent with the speaker classification decision rule. ALVQ aims at providing a compressed, geometrically consistent data representation. It is fit to cover irregular data distributions and computes the distance of the input vector sequence from its nodes. On the other hand, MLP aim at a data representation to fit to discriminate patterns belonging to different classes. It has been shown that MLP nets can approximate Bayesian "optimal" classifiers with high precision, and their output values can be related a-posteriori class probabilities. The different characteristics of these neural models make it possible to devise hybrid neural net systems, consisting of classification modules based on these two different philosophies. The proposed method is compared with LBG algorithm, LVQ algorithm and MLP for performance.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.12 no.3
• /
• pp.250-255
• /
• 2012

Face recognition has wide applications in security and surveillance systems as well as in robot vision and machine interfaces. Conventional challenges in face recognition include pose, illumination, and expression, and face recognition at a distance involves additional challenges because long-distance images are often degraded due to poor focusing and motion blurring. This study investigates the effectiveness of applying photon-counting linear discriminant analysis (Pc-LDA) to face recognition in harsh environments. A related technique, Fisher linear discriminant analysis, has been found to be optimal, but it often suffers from the singularity problem because the number of available training images is generally much smaller than the number of pixels. Pc-LDA, on the other hand, realizes the Fisher criterion in high-dimensional space without any dimensionality reduction. Therefore, it provides more invariant solutions to image recognition under distortion and degradation. Two decision rules are employed: one is based on Euclidean distance; the other, on normalized correlation. In the experiments, the asymptotic equivalence of the photon-counting method to the Fisher method is verified with simulated data. Degraded facial images are employed to demonstrate the robustness of the photon-counting classifier in harsh environments. Four types of blurring point spread functions are applied to the test images in order to simulate long-distance acquisition. The results are compared with those of conventional Eigen face and Fisher face methods. The results indicate that Pc-LDA is better than conventional facial recognition techniques.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.4
• /
• pp.2078-2093
• /
• 2019

Visual smoke recognition is a challenging task due to large variations in shape, texture and color of smoke. To improve performance, we propose a novel smoke recognition method by combining dual-encoded features that are extracted from both spatial and Curvelet domains. A Curvelet transform is used to filter an image to generate fifty sub-images of Curvelet coefficients. Then we extract Local Binary Pattern (LBP) maps from these coefficient maps and aggregate histograms of these LBP maps to produce a histogram map. Afterwards, we encode the histogram map again to generate Dual-encoded Local Binary Patterns (Dual-LBP). Histograms of Dual-LBPs from Curvelet domain and Completed Local Binary Patterns (CLBP) from spatial domain are concatenated to form the feature for smoke recognition. Finally, we adopt Gaussian Kernel Optimization (GKO) algorithm to search the optimal kernel parameters of Support Vector Machine (SVM) for further improvement of classification accuracy. Experimental results demonstrate that our method can extract effective and reasonable features of smoke images, and achieve good classification accuracy.

• Korean Journal of Agricultural Science
• /
• v.50 no.3
• /
• pp.399-406
• /
• 2023

In this study, we propose a ripe tomato bunch localization method based on convolutional neural networks, to be applied in robotic harvesting systems. Tomato images were obtained from a smart greenhouse at the Rural Development Administration (RDA). The sample images for training were extracted based on tomato maturity and resized to 128 × 128 pixels for use in the classification model. The model was constructed based on four-layer convolutional neural networks, and the classes were determined based on stage of maturity, using a Softmax classifier. The localization of the ripe tomato bunch region was indicated on a class activation map. The class activation map could show the approximate location of the tomato bunch but tends to present a local part or a large part of the ripe tomato bunch region, which could lead to poor performance. Therefore, we suggest a recursive method to improve the performance of the model. The classification results indicated that the accuracy, precision, recall, and F1-score were 0.98, 0.87, 0.98, and 0.92, respectively. The localization performance was 0.52, estimated by the Intersection over Union (IoU), and through input recursion, the IoU was improved by 13%. Based on the results, the proposed localization of the ripe tomato bunch area can be incorporated in robotic harvesting systems to establish the optimal harvesting paths.

• Journal of the Korean Society for information Management
• /
• v.26 no.3
• /
• pp.451-462
• /
• 2009

The aim of this study is to apply latent semantic indexing(LSI) techniques for efficient automatic classification of opinionated documents. For the experiments, we collected 1,000 opinionated documents such as reviews and news, with 500 among them labelled as positive documents and the remaining 500 as negative. In this study, sets of content words and sentiment words were extracted using a POS tagger in order to identify the optimal feature set in opinion classification. Findings addressed that it was more effective to employ LSI techniques than using a term indexing method in sentiment classification. The best performance was achieved by a supervised LSI technique.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.3
• /
• pp.297-303
• /
• 2009

In this paper, we propose the biomedical spectral pattern classification techniques by the fusion scheme based on the SpPCA and MLP in extended feature space. A conventional PCA technique for the dimension reduction has the problem that it can't find an optimal transformation matrix if the property of input data is nonlinear. To overcome this drawback, we extract features by the SpPCA technique in extended space which use the local patterns rather than whole patterns. In the classification step, individual classifier based on MLP calculates the similarity of each class for local features. Finally, biomedical spectral patterns is classified by the fusion scheme to effectively combine the individual information. As the simulation results to verify the effectiveness, the proposed method showed more improved classification results than conventional methods.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.2
• /
• pp.55-62
• /
• 2016

Recently, users choosing a restaurant basedon information provided by blogs are increasing significantly. However, those of most blogs are unreliable since domestic restaurant blogs are occupied by advertising postings written by 'power bloggers'. Thus, in order to ensure the reliability of blogs, it is necessary to filter the advertising blogs which are sometimes false or exaggerated. In this paper, we propose the method of distinguishing the advertising blogs utilizing an automatic classification technique. In the proposed technique, we first manually collected advertising restaurant blogs, and then analyzed features which are commonly found in those blogs. Using the extracted features, we determined whether a given blog is advertising one applying automatic classification algorithms. Additionally, we select the features and the algorithm which guarantee optimal classification performance through comparative experiments.