Learning Automata Based Incremental Learning Method for Deep Neural Networks
نویسندگان
چکیده
منابع مشابه
Bayesian Incremental Learning for Deep Neural Networks
In industrial machine learning pipelines, data often arrive in parts. Particularly in the case of deep neural networks, it may be too expensive to train the model from scratch each time, so one would rather use a previously learned model and the new data to improve performance. However, deep neural networks are prone to getting stuck in a suboptimal solution when trained on only new data as com...
متن کاملYuan Liu INCREMENTAL LEARNING IN DEEP NEURAL NETWORKS
TAMPERE UNIVERSITY OF TECHNOLOGY Master’s Degree Program in Information Technology YUAN LIU: Incremental Learning in Deep Neural Networks Master of Science Thesis, 67 pages, 5 Appendix pages June 2015 Major: Pervasive Systems Examiners: Prof. Joni-Kristian Kämäräinen, Dr. Ke Chen
متن کاملA Link Prediction Method Based on Learning Automata in Social Networks
Nowadays, online social networks are considered as one of the most important emerging phenomena of human societies. In these networks, prediction of link by relying on the knowledge existing of the interaction between network actors provides an estimation of the probability of creation of a new relationship in future. A wide range of applications can be found for link prediction such as electro...
متن کاملAnalytical Incremental Learning: Fast Constructive Learning Method for Neural Network
13:20 13:40 Analytical Incremental Learning: Fast Constructive Learning Method for Neural Network Syukron Ishaq Alfarozi, Noor Akhmad Setiawan, Teguh Bharata Adji, Kuntpong Woraratpanya, Kitsuchart Pasupa, Masanori Sugimoto Faculty of Information Technology, King Mongkut’s Institute of Technology Ladkrabang Department of Electrical Engineering and Information Technology, Universitas Gadjah Mada...
متن کاملTowards Incremental Learning with Deep Convolutional Networks
Deep neural networks are a powerful class of machine learning models. However they require a lot of time and computational resources to train. We propose to apply an incremental learning approach to train models by utilizing the information present in pre-trained models. We build towards this goal by studying the relationship between network architecture, categories in training data, the amount...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE Access
سال: 2019
ISSN: 2169-3536
DOI: 10.1109/access.2019.2907645