KMITL

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

Browse

Search Results

Now showing 1 - 7 of 7

Some of the metrics are blocked by your
consent settings
Monitoring Pesticide Residue on Bok Choi using Convolution Neural Network with NIR spectral Data
(2024-12-29)
Lapcharoensuk, Ravipat
;
Fhaykamta, Chawisa
;
Anurak, Watcharaporn
;
Chadwut, Wasita
Deep learning (DL) has been applied in agriculture, especially quality control in agricultural processing. One key area of interest is the detection and monitoring of pesticide residues in crops. The most popular measurement tool for nondestructive monitoring of pesticide residues is near-infrared spectroscopy (NIRS). A combination of CNN model with NIR spectral data was developed for monitoring pesticide residue on bok choi. The NIR spectral of bok choi with and without pesticide residue (chlorpyrifos) was collected in wavelength range between 908 and 1676 nm. A simple structure of CNN was modified for a one-dimensional task and this deep learning architecture was trained for classification of the bok choi samples. The results showed prefect prediction with 100% accuracy, precision, recall and specificity. This study also found that deep learning for NIR spectroscopy data requires less processing than traditional machine learning while still achieving great results.
Some of the metrics are blocked by your
consent settings
Fake News Detection on Social Media: Case Study of 2019 Novel Coronavirus
(2021-12-17)
Kowirat, Rutchaneewan
;
Boongasame, Laor
Fake news is news that is created with the intent to deceive and mislead readers. It is a problem that occurs in every era because it creates misunderstandings for people through a variety of media channels such as newspapers, radio, or television. Nowadays, fake news has become a big problem. When social media has become another channel to increase the spread of fake news and came to play a big role during the epidemic like COVID-19. Fake news creates panic and creates false knowledge of how to protect yourself from COVID-19. Therefore, the objective of this research is to create a method that can detect fake news on social media. It focuses only on news related to COVID-19. In addition, the information was extracted directly from social media such as Twitter. Moreover, this research applying machine learning processes to classify real and fake news. From the experimental results, the accuracy was measured at 99.92% with the Decision Tree model.
Some of the metrics are blocked by your
consent settings
Building Minimal Classification Rules for Breast Cancer Diagnosis
(2018-08-06)
Douangnoulack, Phonethep
;
Boonjing, Veera
A rule based classifier is widely applied in breast cancer diagnosis. The classifier with a good performance of disease classification have been developed and highly required over the past decades. Since classification rules are derived from previous diagnosis with a large amount of features, it challenges to build a minimal number of rules with high performance while retaining all diagnosis information. The Principal Component Analysis (PCA) is known as a lossless data reduction technique with good classification performance. Therefore, this paper aims at finding the best performance classifier giving minimal classification rules by employing PCA. Based on experiment result on Wisconsin Breast Cancer data set, the J48 decision tree classifier is found to be the best among the three classifiers: J48 decision tree, Reduced Error Pruning Tree, and Random Tree.
Some of the metrics are blocked by your
consent settings
Classifying attitudes of thai business organizations toward the perceived benefit of customer predictive analytics
(2018-07-02)
Limpeeticharoenchot, Santisook
;
Cooharojananone, Nagul
;
Chavarnakul, Thira
;
Atchariyachanvanich, Kanokwan
In this paper, we applied K-means++, Agglomerative and Decision Tree techniques to classify characteristics of Thai business organizations toward the perceived benefit of predictive analytics. We believe that different characteristics of business organizations require different solution approaches to overcome challenges and gain business value. In the exploratory survey, we collected 84 respondents, who are data analysts, business analysts, business owners, management, and IT managers. The respondents' size of organization varies from small firms to large firms. The result from three techniques of classification shows that Thai business organizations perceiving the benefit of predictive analytics could be divided into two groups. We further confirmed that with the analysis of variance technique to identify the difference of means in each parameter. We also found that the most important descriptive profile is led by customer-related factors such as a change in the percentage of customers, the number of direct customers, and number of sales and marketing staff. These are followed by technology factors, which are the number of external and internal data sources and number of analytics and technologies currently used for data analytics.
Some of the metrics are blocked by your
consent settings
Bank direct marketing analysis of asymmetric information based on machine learning
(2015-08-24)
Ruangthong, Pumitara
;
Jaiyen, Saichon
The bank direct marketing campaign for offering products that meet the customers' needs is the challenge problems. The bank direct marketing data analysis is important work that helps the banks predict whether customers will sign a long term deposits with the banks. The method that can predict such customers' needs can be profitable to the banks for improving their marketing campaign strategies. Unfortunately, it is very hard to predict the customers' needs because the available information is asymmetric. In this paper, we propose the method to analyze asymmetric information using SMOTE algorithm and Rotation Forest (PCA)-J48. The SMOTE method is used to modify the data and improve the accuracy of the prediction. The performance of the proposed method is evaluated and compared to Decision Tree, Rotation Forest, Navie Bayes, BayesNet, Multilayer Perceptron Neural Network, RBF Neural Network. The experimental results show the predicting accuracies of all predictors. The experiments show that Rotation Forest (PCA)-J48 can achieve the highest value of accuracy and specificity. However, the sensitivity of Rotation Forest (PCA)-J48 is higher than all methods except BayesNet and Rotation Forest (PCA) RandomTree.
Some of the metrics are blocked by your
consent settings
A hybrid ensemble of machine and statistical learning using confidence-based boosting
(2015-01-01)
Chairatanasongporn, Nattawut
;
Jaiyen, Saichon
Nowadays, the classification problems have become more challenging due to the various types of data set. Some data are appropriated for machine learning techniques and some data are appropriated for statistical leaning techniques. This work proposes a new hybrid ensemble of machine and statistical learning models using confidence-based boosting. The proposed method which uses variants of based classifiers can solve classification problems in variant data set. Moreover, combining the confidence value to the current boosting method can improve the performance of classification. The performance of proposed method is compared to the ensemble of decision trees and MRN created by Adaboost.M1 on data sets from UCI. The experimental results show that the proposed method can improve the accuracy in both binary and multiclass classification problems.
Some of the metrics are blocked by your
consent settings
Association rules for data mining in item classification algorithm: Web service approach
(2012-07-16)
Phankokkruad, Manop
The student's assessment is the core of learning process, which facilitates teachers to evaluate a student's knowledge level; furthermore, the precise measurement helps the students knowledge development reaches their full potential. Usually, this assessment method is also known as computer adaptive testing (CAT). The conventional CAT systems contain its own item bank, which is stored separately in many repositories over the Internet. The collection of the items from many repositories of database together makes these items were reused, sharable, valuable, and also makes the larger item bank. Unfortunately, the combined items make the tangled data, and greater data size. The problem of data overloaded occurs, and a large number of irrelevant and redundant data should be eliminated. This paper has attempted to formulate the data mining model in manipulate the optimal item-set from the different sources of the item. The item data from many repositories were mined in order to extract the implicit, useful information and interesting patterns from the huge irrelevant and redundant data collections. Therefore, the association rules were established by applying the knowledge pattern, decision trees, adaptive testing and related theory. The result shows that the association rules and mining process are used to create the optimal item-set. This optimal item-set was delivered through Web service to the CAT applications. The result also shows that data mining works properly. Moreover, the precise items help the students improve their knowledge reach their full potential. © 2012 IEEE.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results