Using rough set through for classification of image segmentation data

• Main Author: Diallo Abdoulaye, Kindy
• Format: Thesis
• Language: English
• Published: 2014
• Subjects: QA Mathematics
• Online Access: http://eprints.utm.my/id/eprint/48667/1/DialloAbdoulayeKindyMAIS2014.pdf; http://eprints.utm.my/id/eprint/48667/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:82003?queryType=vitalDismax&query=Using+rough+set+through+for+classification+of+image+segmentation+data&public=true
• Institution: Universiti Teknologi Malaysia

Knowledge Discovery in Database (KDD) can be defined as a technology or a process that helps to extract valuable information including hidden and unseen patterns, trends and relationships between variables from a large amount of data. The information learnt and the discovery made can help in applying the new found pattern in the training set to an unseen data, known as test set, that can guide and facilitate a crucial business decision making task. A large number of data mining techniques have been proposed in the literature for classification purpose. In this work, we are using the Rough Set Classifier (RSC) for mining image segmentation data set obtained from an online machine learning data repository. The RSC is a rule based data mining technique which generates rules from large databases and has great capabilities to deal with noise and uncertainty in data set. In order to find out the best accuracy method, we conducted around 10 experiments by varying the proportions between the training and test sets. The best method gave us an accuracy of 85.71%.