SIG Learning

SIG Chair: Yuh-Jye Lee (Professor, Dept. of Applied Mathematics, National Chiao Tung University)
Four sub-projects pertain to SIG Learning:

 
The Aims and Objectives of SIG Learning: 
In SIG-Learning, so far, we have three major projects. All of them follow their schedule described in the RPD very well. We will take the comments and feedbacks from the steering committee. Meanwhile, we will pay our attention on the applications in smart factory.
Interactive machine learning project has designed the general optimization formulation for active learning that aims to work on the interactive learning scenarios. For the active learning strategy, we have developed a unify algorithm that considers uncertainty, diversity and representativity for data annotation. More generally, we can discuss the scenario where more than a human (oracle) candidate can be consulted for the annotation tasks in a humans-machines integrated unit. Among different humans/oracles, we have various types such as between enthusiastic or reluctant, different availability, different data correctness, different costs, to name a few.
In the project, Interactive Dialogue Learning, they will keep developing methods for characterizing data-driven dialogue systems. The Internet movie script database (IMSDB) will be used in developing the characterized dialogue system and conduct experiments on it. We will collaborate with the Intel champions in this part very deeply. Also, keep developing a DRL-based interactive question answering system. We expect that questions and answers could be generated automatically from the given documents. Based on the discussion with the Delta champions, this technique will fit their needs.
In Learning in Compressed Domain part, the power of compressed learning has been demonstrated in ECG dataset. They will try to test and evaluate the proposed learning model. Also develop the aggregation strategy for combining diversified ELM.
Finally, we will try to expand our SIG Learning. We have invited Prof. Sun Min from EE, NTHU to join our SIG. Prof. Sun has submitted his RPD entitled Learning to Act from Demonstration. We are looking forward to his contribution and believe that this will be an important part of our SIG as well as ACB main project. (updated in Feb, 2017)