Deep learning for image processing in WEKE environment

• Main Authors: Zainudin, Z., Shamsuddin, S. M., Hasan, S.
• Format: Article
• Language: English
• Published: International Center for Scientific Research and Studies 2019
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://eprints.utm.my/id/eprint/90130/1/ZanariahZainudin2019_DeepLearningforImageProcessing.pdf; http://eprints.utm.my/id/eprint/90130/; http://home.ijasca.com/data/documents/1_page1-21_Deep-Learning-for-Image-Processing-in-WEKA-Environment_1.pdf.
• Institution: Universiti Teknologi Malaysia

Deep learning is a new term that is recently popular among researchers when dealing with big data such as images, texts, voices and other types of data. Deep learning has become a popular algorithm for image processing since the last few years due to its better performance in visualizing and classifying images. Nowadays, most of the image datasets are becoming larger in terms of size and variety of the images that can lead to misclassification due to human eyes. This problem can be handled by using deep learning compared to other machine learning algorithms. There are many open sources of deep learning tools available and Waikato Environment for Knowledge Analysis (WEKA) is one of the sources which has deep learning package to conduct image classification, which is known as WEKA DeepLearning4j. In this paper, we demonstrate the systematic methodology of using WEKA DeepLearning4j for image classification on larger datasets. We hope this paper could provide better guidance in exploring WEKA deep learning for image classification.