Ģ��ֱ��

Every day we face situations where we must balance between multiple objectives to try to find the best compromise. Such situation could be as simple as choosing what to eat for breakfast or they could be as complicated as designing policies on a national scale with billions of Euros hanging in the balance. In his doctoral thesis Saini Bhupinder has created an algorithm-based solution to help solve multiple objectives in an effective manner.

Companies and organizations often find themselves in a situation where they have several objectives to pursue. In such a situation, compromises have to be made in order to achieve the objectives. Often it may not be possible to fully meet all of them.

In his thesis Bhupinder Saini created multiple algorithms to solve problems with multiple objectives in an efficient manner. The goal is to save both time and money in the decision-making progress while helping the decision-maker to control the process of problem solving.

“It is important for the decision-maker to understand the process behind the algorithms. That way they can trust the solutions proposed by the algorithm. This is why I also created a new style of visualization to help to understand the compromises and what they mean in order to make better choices.”, Saini explains.

While using the DESDEO framework, a software created by Saini and his colleagues, the decision-maker can see how compromises affect the outcome. The user can balance between different solutions to find the best outcome while understanding how different compromises affect the possible outcomes.

In his thesis the algorithms were used to solve a problem from metallurgical engineering designing new alloys of steel while balancing 6 objectives. He helped two decision-makers arrive at a fruitful compromise and find new and strong alloys of steel.

Free software is used by students and researchers

When researchers create new algorithms, they often do not make them easily accessible. DESDEO, created by Saini and his collegues, is however available for anyone to use freely at desdeo.it.jyu.fi.

“Students use DESDEO to learn about the algorithms mentioned earlier, and decision-makers use it to solve their problems. DESDEO is also open to contributions from researchers from all over the world. These researchers can use the tools provided within DESDEO to create their own new algorithms quickly and easily.”, Saini says.

The dissertation has been published in the Ģ��ֱ�� Dissertations series JYU Dissertations, number 556, Jyväskylä 2022. ISBN 978-951-39-9196-8 (PDF), URN:ISBN:978-951-39-9196-8, ISSN 2489-9003. Permanent link to this publication:

M.Tech. Bhupinder Saini defends his doctoral dissertation in Computational Science "Pioneering Techniques to Tackle Challenges of Interactive Multiobjective Optimization". Opponent Professor Jyrki Wallenius (Aalto University) and Custos Professor Kaisa Miettinen (Ģ��ֱ��). The doctoral dissertation is held in English.

The audience can follow the dissertation in the lecture hall (S212) or online. Link to the online event: