Seminar: A Sequential Deep Learning Algorithm for Robust Mixed-integer Optimisation Problems, 12th May, 1PM

When: Thursday 12th of May, 1PM AEDT

Where: The talk will be presented online via Zoom, RSVP here.

Speaker: Mohammadreza Chamanbaz

Title: A Sequential Deep Learning Algorithm for Robust Mixed-integer Optimisation Problems

Abstract:

Optimisation has long been a key driving force in solving machine learning (ML) problems. However, there have been a number of very recent attempts to use machine learning to assist in solving continuous and mixed-integer optimisation problems. This talk uses a neural network classifier to considerably reduce the computational complexity of solving robust optimisation problems. To this end, we first approximate the robust problem with a sampled optimisation problem and use a sequential algorithm to solve the sampled optimisation problem. A sequential algorithm is proposed that uses the neural network classifier to identify the subset of constraints constructing a candidate solution. We then evaluate the algorithms’ effectiveness on a variety of computational examples.

Bio:

Mohammadreza Chamanbaz received his PhD in Control Science from the Department of Electrical & Computer Engineering, National University of Singapore, in 2014. He was with Data Storage Institute, Singapore as a research scholar from 2010 to 2014 and with Singapore University of Technology and Design as a postdoctoral research fellow from 2014 to 2017. Dr Chamanbaz was an assistant Professor at Arak University of Technology from 2017 to 2019 and a senior research fellow at iTrust Centre for Research in Cyber-security from 2019 to 2020. He is currently a research fellow at Australian Centre for Field Robotics, and his research activities are mainly focused on probabilistic and randomised algorithms for solving uncertain optimisation problems.

Contacts

Sydney Institute for Robotics and Intelligent Systems
info@acfr.usyd.edu.au