- Regarding ICCAS 2016
- History
- Committee
- Call for Paper
- Sponsor
- Submission
- Paper Submission Guideline
- Online Paper Submission
- List of Organized Session
- Submission of OS Proposal
- Program
- Program at a Glance
- Plenary Session
- Workshop & Tutorial
- Technical Tour
- Special Session
- Session Timetable
- Technical Sessions
- List of Digest
- Exhibition
- Social Program
- Registration
- Guideline
- Online Registration
- Author's Kit
- Awards
- Guidelines for Lecture Session
- Guidelines for Interactive
Poster Session - Travel Information
- General Information
- Transportation
- Accommodation
- Venue
- VISA Information
- Tour Information
Workshop & TutorialICCAS2016 > Program > Workshop & Tutorial
The conference program will include Technical Workshops & Tutorials on Sunday, October 16.
The workshops and tutorials of this year will be specially interesting to those who want to learn about the latest trends and subjects.
Registration Process: Log in and then go to Mypage > 'Tutorial & Technical Tour' menu
Registration Deadline: September 30, 2016 It is possible to pay tutorial registration fee on site.
Registration Deadline: September 30, 2016 It is possible to pay tutorial registration fee on site.
| Tutorial | Machine Learning with Sequential Data |
|---|---|
| Time & Place | 13:00~17:00, Room 105 |
| Organizer | Prof. Jaesik Choi and Prof. Jun Moon (Ulsan National Institute of Science and Technology) |
| Fee | Student 50,000Won, Regular 100,000Won |
| Presentation | English |
| Program | Handling sequential data is important for various applications including localization of airplane and autonomous car, tracking objects in surveillance missions, and predicting the changes of financial markets. In this tutorial, we will overview various methods to model, inference with and learn from sequential data. First, we will present the basic principles of dynamical systems including Hidden Markov Models, Linear Dynamical Systems and Nonlinear Dynamical Systems. It will cover the basic models, state estimation and parameter learning algorithms. Second, we will present the basics of sequential decision processes including Linear-Quadratic-Gaussian, Markov Decision Processes and Reinforcement Learning. Then, we will introduce Nonparametric Bayesian learning algorithms for sequential data analysis. This section will introduce Gaussian Processes, Generalized Wishart Processes, Dirichlet Processes and its recent applications, the Automatic Statistician System, which generated human-readable report automatically from time-series data. Finally, we will overview recent advances in deep learning for sequential data. This section will include Recurrent Neural Networks, Long short-term memory, and Neural Turing Machine. The target audience of this tutorial is junior graduate students in engineering field. We will expect basic knowledge for linear algebra and statistics. However, prior knowledge for machine learning and control is not required. 1. Basics of Sequential State Estimation - Hidden Markov Models (Learning, forward-backward algorithm) - Linear Dynamical Systems (Inference and Learning with Kalman filter) - Nonlinear Dynamical Systems (Inference and Learning with Extend KF and Particle Filter) 2. Basics of Sequential Decision - Linear-Quadratic-Gaussian (LQG) control - Markov Decision Process - Reinforcement Learning 3. Nonparametric Bayesian for Sequential Data - Gaussian Processes - Generalized Wishart Processes - Dirichlet Processes - The Automatic Statistician System for financial data analysis 4. Deep Learning with Sequential Data - Deep Learning - Recurrent Neural Networks and its application in EEG data - Long short-term memory - Neural Turing Machine |
| Tutorial | |
|---|---|
| Time | 10:00-17:30 |
| Organizer | Prof. Hamid Reza Karimi (University of Agder) |
| Fee | Student 150 USD, Regular 200 USD |
| Presentation | English |
| Program | Markovian jump systems have been widely used to describe many practical systems, such as fault-tolerant systems, communication systems, power systems, economics systems and so on. The main aim of this tutorial is to provide an introduction to the basic principles and applications of Markovian jump systems in control systems and practice. In this workshop, we will deliver highly useful knowledge and experience for graduate students, scientists, and field engineers interested in this research area.
The workshop will begin with an introduction to the state-of-the-art of Markovian jump systems and present main challenges and more recent developments and progresses in this context. Then, conventional stability analysis of these systems will be presented with some discussions on more complex Markovin jump systems including singular and time delay systems. In the sequel, control synthesis in the form of state feedback and output feedback controllers are presented for Markovian jump systems and some recent developments are presented and discussed. Then, the problems of observer designs for these systems are presented, followed by some special and important cases of Markovian jump systems in practice. Finally, some practical examples will be presented to demonstrate the use of Markovian jump systems in practice.
|
| Tutorial | |
|---|---|
| Time | 13:00~17:00 |
| Organizer | Prof. Chul-Goo Kang (Konkuk University) |
| Fee | Student 50,000Won, Regular 100,000Won |
| Presentation | If there are foreigners, it will be presented in English. If all audience are Korean, it will be in Korean. |
| Program | In this Tutorial, Input Shaping ® control, a kind of feedforward control, is introduced from basis to application, which can suppress residual vibrations via manipulating reference inputs appropriately. We present a simple technique of input shaping, and discuss advantages and disadvantages of this technique for engineers of industry, and researchers/students of research institutes and universities. Specifically, we discuss
|
