Predictor based on Least Squares
Description
Nowadays, model development for output quality prediction purpose is highly required for many process industry applications. A study of applying Least Square Support Vector Machine modelling technique to develop an output quality of synthetic polymer production has been done with a promising result. In addition, the developed model performance has been compared with some Multivariate Statistical Process Control (MSPC) modeling techniques, such as Partial Least Square, Principal Component Analysis combined (PCA) with LSSVM, etc. Modeling work done during the study will be discussed in the lecture including an overview of LSSVM technique. The lecture will not discuss many details on theory and mathematical part, but more on the modeling technique overview and its application(s).
Speaker(s)
Eko Harsono MEng
Prolab
Location
B 1.12, HAN Faculty of Engineering
Ruitenberglaan 26, 6826 CC ARNHEM
Organiser
Measurement, Control and Steering Technology
HAN Master of Control Systems Eng.
Name and contact details for information
Tanja Vermeulen (026-3658156)