News

titled Decoding control: scalable MPC approximation with decision trees.

titled When Global Optimization Meets Machine Learning: Challenges and Opportunities.

Our recent work was published in the Proceedings of the International Conference on Machine Learning (ICML). In this work, we proposed a new approach combining bound tightening and reduced space branch and bound, which has shown to solve k-center clustering problems with up to 10 million samples in serial.

The paper has been cited for more than 500 times.

Kaixun’s recent work was published in the Proceedings of the International Conference on Machine Learning (ICML). In this work, we proposed a new global optimization algorithm for clustering problems and demonstrated that our algorithm can solve the MSSC problem with up to 200,000 samples in parallel, while state-of-the-art approaches in the literature can only deal with two thousand samples.

Mahdi’s recent work was published on IEEE Transactions on Power Systems, in which he proposed a new global solver for Transmission Expansion Planning with AC network (ACTEP).