KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 63

Some of the metrics are blocked by your
consent settings
Comparison of Sampling Methods for Imbalanced Data Classification in Random Forest
(2019-01-10)
Paing, May Phu
;
Pintavirooj, C.
;
Tungjitkusolmun, Supan
;
Choomchuay, Somsak
;
Hamamoto, Kazuhiko
Imbalanced data classification is a serious and challenging task for most of the medical image diagnosis applications. They usually produce a larger number of false samples compared to the actual ones. That is the number of samples for the class of interest (minority) is significantly fewer than other types of class (majority). The classification performed using such data is called imbalanced data classification. As a consequence, the learning model bias towards the majority class and fails the classification of the minority class. Data sampling and ensemble methods are common ways to compensate for this issue. Random forest (RF), an ensemble of multiple decision trees, is very famous in both of the classification and regression problems because of its robust and accurate predictions. However, it also suffers class bias in the imbalanced data classification problems. This paper proposes and compares different sampling methods to solve the imbalanced data classification in RF.
Some of the metrics are blocked by your
consent settings
Suitable Supervised Machine Learning Techniques for Malignant Mesothelioma Diagnosis
(2019-01-10)
Win, Khin Yadanar
;
Maneerat, Noppadol
;
Choomchuay, Somsak
;
Sreng, Syna
;
Hamamoto, Kazuhiko
Malignant Mesothelioma (MM) is a rare, aggressive cancer that grows in the lining of the internal organs such as lung, abdomen or heart. Fousing on MM diagnosis, in this paper, we investigate multiple machine learning methods and compare for accurate MM diagnosis results. Seven machine learning algorithms namely (i) Linear Discriminant Analysis (LDA), (ii) Naïve Bayes, (iii) K Nearest Neighborhood (KNN), (iv) Support Vector Machine (SVM), (v) Decision Tree (DT), (vi) Logistic Regression (LogR) and (vii) Random forest (RF) algorithms are exploited. The experiments dataset containing 324 cases with 34 features and six performance measures are used to assess the accuracy of evaluated classifiers. The average accuracy of LDA, NB, KNN, SVM, DT, LogR and RF are 61.73%, 67.90%, 91.36%, 100%, 100%, 100% and 100%, respectively. In addition, the computational complexity of each method is also analyzed. Each algoritm is judged based on its classification accuracy and computational complexity. It is found that SVM, DT, LogR and RF outperform the others and even previous studies.
Some of the metrics are blocked by your
consent settings
An Investigation of UWB Path Loss for Body Area Network
(2018-12-24)
Sansoda, Buncha
;
Choomchuay, Somsak
This paper presents the investigation of body area communication channel and provide path loss evaluation by measurement compared with channel model of IEEE 802.15.6 standard. The measurement performs in ultra-wideband frequency range and setup the body in different body poster. The receive antenna is mounted on waist while transmit antenna is placed on wrist and ankle. We found trend of path loss similar as model. The results shows that transmission between the two antenna are highly attenuated when the radio wave propagates across the body. However, if the wave direction propagates along the body as waist to ankle, the path loss would slowly increase. Thus the suitable position of antenna would certainly be important with measurement result to reduce the body effect.
Some of the metrics are blocked by your
consent settings
Detection and classification of overlapping cell nuclei in cytology effusion images using a double-strategy random forest
(2018-09-11)
Win, Khin Yadanar
;
Choomchuay, Somsak
;
Hamamoto, Kazuhiko
;
Raveesunthornkiat, Manasanan
Due to the close resemblance between overlapping and cancerous nuclei, the misinterpretation of overlapping nuclei can affect the final decision of cancer cell detection. Thus, it is essential to detect overlapping nuclei and distinguish them from single ones for subsequent quantitative analyses. This paper presents a method for the automated detection and classification of overlapping nuclei from single nuclei appearing in cytology pleural effusion (CPE) images. The proposed system is comprised of three steps: nuclei candidate extraction, dominant feature extraction, and classification of single and overlapping nuclei. A maximum entropy thresholding method complemented by image enhancement and post-processing was employed for nuclei candidate extraction. For feature extraction, a new combination of 16 geometrical and 10 textural features was extracted from each nucleus region. A double-strategy random forest was performed as an ensemble feature selector to select the most relevant features, and an ensemble classifier to differentiate between overlapping nuclei and single ones using selected features. The proposed method was evaluated on 4000 nuclei from CPE images using various performance metrics. The results were 96.6% sensitivity, 98.7% specificity, 92.7% precision, 94.6% F1 score, 98.4% accuracy, 97.6% G-mean, and 99% area under curve. The computation time required to run the entire algorithm was just 5.17 s. The experiment results demonstrate that the proposed algorithm yields a superior performance to previous studies and other classifiers. The proposed algorithm can serve as a new supportive tool in the automated diagnosis of cancer cells from cytology images.
Some of the metrics are blocked by your
consent settings
License Plate Detection of Myanmar Vehicle Images from Dissimilar Angle Conditions
(2018-08-21)
Khin, Ohnmar
;
Phothisonothai, Montri
;
Choomchuay, Somsak
It has been studied that there is no established LPR (License Plate Recognition) to detect and identify the license plates from dissimilar angles. The aim of the paper is to detect the dissimilar angles of the license plate with the non-fixed LPR system. Therefore, the horizontal and vertical dilation, skew angle detection and automatic bounding box have been proposed to detect the license plate. The proposed method has been applied to the four different types of Myanmar license plates, e.g., private cars, taxi, tour buses and religion cars. One car each is taken into four different types of angles on the dissimilar conditions. Experimental result indicated that this method can detect the disparate types of license plates with a high accuracy, i.e., the proposed approach achieved a favorable outcome rate of 97% at 100 license plates.
Some of the metrics are blocked by your
consent settings
Improved Random Forest (RF) classifier for imbalanced classification of lung nodules
(2018-08-13)
Paing, May Phu
;
Choomchuay, Somsak
Computer-aided detection (CAD) for lung cancer acts a dynamic research in biomedical engineering. These CADs generally occur imbalanced data classification because there are a large number of false lesions which are non-nodules (majority class), compared to the actual nodules (minority class). This paper proposes an improved random forest (RF) classifier to solve the learning bias problem of the imbalanced classification. The proposed RF applies sampling and three feature selection schemes namely Relief, Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to improve the classification performance.
Some of the metrics are blocked by your
consent settings
Detection and Recognition of Myanmar Characters from the Dissimilar Images
(2018-08-06)
Khin, Ohnmar
;
Phothisonothai, Montri
;
Choomchuay, Somsak
The impartial of this paper is to spread the License Plate Recognition for the traffic light development of Myanmar. In this paper, image processing toolbox, controlling and take boxes were proposed to extract the numbers and characters from the input images. Our method consists of three steps. In the first step, the input license plate image is increased by doing some the action or noise modification. Then the features of the characters are extracted to overcome the sameness problems of the Myanmar characters. Finally, neural network is used for the character recognition. The experiments were carried out on the dissimilar Myanmar vehicle images and results showed the efficiency of 93% which was substantially applied for the license plate detection, letters and numbers extraction.
Some of the metrics are blocked by your
consent settings
License Plate Detection of Myanmar Vehicle Images Captured from the Dissimilar Environmental Conditions
(2018-06-21)
Khin, Ohnmar
;
Phothisonothai, Montri
;
Choomchuay, Somsak
Due to the difficulty in defecting the images, in this paper, we proposed the method that can detect region of interest of the license plate for Myanmar vehicles captured from the dissimilar environmental conditions, e.g., different type of license plates, angle of image capturing, and real environmental conditions. In this paper, the horizontal and vertical dilation, skew angle detec-tion and automatic bounding box were proposed to detect the license number from input images. In our experiment, the ob-tained results showed that an average detection accuracy of 99% which was substantially applied for the license plate detection from the dissimilar environmental conditions.
Some of the metrics are blocked by your
consent settings
A Study of Pulmonary Nodules from Multi-slice Computed Tomography Using 3-D Structure
(2018-06-21)
Paing, May Phu
;
Choomchuay, Somsak
Multi-slice CT screening is an auspicious imaging test to detect the lung cancer. Formation of the pulmonary nodules in the lungs is the first suspicious sign of the lung cancer. The nodules appear as round or irregular shaped spots having size up to 30 mm on the CT scan. As the rate of cancer cases increases day by day, the manual detection of pulmonary nodules becomes a struggling way for the radiologists. The proposed method aims to develop an automated system to detect the pulmonary nodules using image processing techniques. Moreover, the needs of the conventional 2D detection are figured out and compensated by creating 3D structure.
Some of the metrics are blocked by your
consent settings
Myanmar character extraction from vehicle images using aspect ratio and bounding box
(2018-05-30)
Khin, Ohnmar
;
Phothisonothai, Montri
;
Choomchuay, Somsak
In this paper, we publicize Myanmar character extraction system using the license plate number as experimental sample. In recent times, the identification of car license becomes a popular task because of the increase in the number of vehicles. To settle this problem, there are a number of techniques in which aspect ratio and bounding box are the suitable technology. We also use edge boxes in order to separate the background from the foreground and accentuate in the foreground. In the investigation, the Myanmar character are segmented firstly and then these are extracted precisely. The experimental results showed that the characters have been correctly extracted by the accuracy rate of 90%.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results