Publications by authors named "Muhammad Zafran Muhammad Zaly Shah"
Sensors (Basel)
April 2022
Article Synopsis
- Data streaming applications, especially in IoT, often deal with unlabeled sequential data from sensors, making traditional supervised learning challenging.
- The online manifold regularization technique helps with learning from partially labeled data but typically requires the determination of the radial basis function (RBF) kernel width parameter, which is difficult to set without a lot of labeled data.
- The proposed solution eliminates the RBF kernel by integrating prototype learning, allowing for faster learning and better classification performance, making the approach more practical for scenarios with limited labeled data.
