Optimization Community | Perth, Australia
August 1, 2024

Iain Jardin
Regional Sales Director, APAC, Gurobi Optimization

Generative AI is the new hot topic, and here at Gurobi, we are keeping a close eye on its possible uses in our industry and a big ear out to hear how our users are using it to extract value for their businesses. In this talk, we will briefly introduce Generative AI, particularly LLMs and the Transformer model, discuss areas where this tool can be useful in the field of Decision Intelligence, and provide an update on the current viewpoints here at Gurobi. Also, we would love to hear what our users are currently doing with Generative AI at this session.

Steven Edwards
Team Lead – Technical Account Management, APAC, Gurobi Optimization

In industry and academia, we develop algorithms to meet particular goals. How can we be sure the algorithm is correct? And how can we demonstrate to the end user that the algorithm will work for them? In this presentation, we will discuss the methods we use to verify algorithms using Gurobi and behaviour tests.

Dr Christina Burt
Principal Optimisation Specialist, Fortescue

Many industrial problems require us to repeatedly solve the same optimisation problem albeit with different input data. The parameters of our optimiser can influence things like runtime. Hyper parameter tuning has been known as a way of stabilising these parameters, however, this traditionally requires a large static training set. In this talk we present an “online” approach, which combines the contextual bandit paradigm with our optimiser to learn the best hyper parameter settings on the fly.

Dr Gaurav Singh
Founder, Decision 360

Brendan van Rooyen
Decision Scientist, Decision 360 

By 2050, the electrical grid will be a fundamentally different system to the one we have today. Over the next 20 years, we need to learn how to plan, build, and run a new type of grid while preventing it from going down, even for a second.

Undertaking such a monumental exercise is only possible with mathematical optimisation. Modelling allows us to simulate and understand the future grid before we build it. We’re able to understand the requirements and costs of building and operating the grid over the coming decades. Furthermore, we can use modelling to explore the risk of events like blackouts, big price increases, and the impact of delayed transmission and power station build.

Oliver Nunn
Director, Endgame Economics