KISTI Presents Traffic Congestion Prediction in Major Cities Through
Intelligent Information Technology Learning Simulation Based on MATLAB,
Planning to Share Diverse Customer Cases

MathWorks (kr.mathworks.com), a technical computing software company, announced that it will hold 'MATLAB EXPO 2016 Korea' on April 28 at the COEX InterContinental Hotel in Samseong-dong, Seoul.

MATLAB EXPO is the largest comprehensive technical conference held annually by MathWorks Korea, providing engineers and researchers from various industries with opportunities to learn about the latest technology updates and related solutions for MATLAB and Simulink product families, and to exchange knowledge.

At this event, Chris Hayhurst, Director of MathWorks Consulting Services, will deliver the keynote address covering new features and various use cases of MATLAB and Simulink applied to engineering data analysis. The customer keynote address will be presented by Professor Seon Woo Myung-ho from the Department of Future Automotive Engineering at Hanyang University.


▲The conference will introduce the application of MATLAB and Simulink in various fields including big data analytics, Internet of Things, humanoid robots, and smart cars.

In addition, more than 20 technical sessions will be conducted across five tracks: △Signal and Image Processing, △Language of Technical Computing, △Control Design Automation, and △Tech Talk Special 1 and 2. Participants can experience everything about MATLAB and Simulink product families in the respective sessions and gain knowledge that can be directly applied to work and research, along with customer success cases and solutions. Additionally, at demo booths from various partners, attendees can directly experience the latest technological trends in big data analytics, Internet of Things, humanoid robots, smart cars, and more.

Notably, at this event, Dr. Lee Hong-seok from the Korea Institute of Science and Technology Information (KISTI) will present on predicting traffic flow through artificial intelligence-based analysis of GPS location data in major cities (vehicle data averaging 230,000 vehicles per day in Seoul). The backpropagation algorithm, which was traditionally used mainly for artificial neural network training, required long learning time and failed to properly support multi-layer artificial neural network structures, making it difficult to extract advanced-level knowledge from machine learning data.

To address this, KISTI applied MATLAB's parallel programming performance, which can process large-scale computations quickly and accurately, and successfully predicted real-time traffic conditions through learning simulation of the latest deep learning algorithm RNN (Recurrent Neural Network). Thereafter, KISTI plans to develop a supercomputing-based deep learning algorithm using MATLAB for large-scale nationwide traffic flow prediction in Korea.

On the day of the event, Hanwha Techwin and S&T Heavy Industries will also present cases of applying MATLAB and Simulink to robot and industrial control platform development and commercial vehicle software design, respectively.

For detailed information and pre-registration for MATLAB EXPO 2016, please visit the official MATLAB EXPO website (www.matlabexpo.com/kr).