We study the problem of deploying cybersecurity mitigations to delay the completion times of a set of attacker projects. Existing literature studies how to select a portfolio of mitigations subject to a budget constraint to optimally delay such attack projects, but lacks consideration of the time it takes to implement the mitigations. We assume attackers are working on their projects at the same time the defender is working to complete mitigations to delay those attacks. Our proposed model captures the importance of scheduling mitigations, since to be useful they must be completed before an attacker exploits a vulnerability. This paper introduces an integer programming model of this problem that combines a defender's mitigation scheduling problem with the attackers' project completion problems. We consider alternative solution methods and examine the benefit of the proposed model.

Researchers around the globe participate to the International Symposium on Mathematical Programming to set forth their latest results in mathematics, algorithms, computation, and modeling. As of March 2024, the 2024 edition held in Montréal, has around 1000 participants registered and includes around 600 talks. Building a reasonable schedule with these figures cannot be done by hand. One has to wonder: should the assignment of its talks using mathematical programming be pipelined?

The assignment of the ISMP's presentations is modeled using an enumeration of promising aggregations of presentations which are formed by pre-processing common keywords. A bonus scheme is used to induce an order on the assignation of the aggregations for a given session. The suggested streams from the participant's submissions are mined to select possible streams for given keywords. The modeled problem is solved using the Gurobi Optimizer.

Inherent constraints of conferences are considered such as unallowed parallel presentations for the same speaker, session chair assignation, participant's time slot unavailabilities, and session's filling requirements.

Finally, a re-optimization process is designed to ensure that participant constraints communicated after the publication of the first schedule affect the least amount of assignations.

In this study, we consider a new scheduling problem which focuses on a two-machine reentrant open shop with no-wait and exact delays constraints, where each job should go back to the first machine, i.e. reentrance, to execute a second task after a fixed time lag that separates the two operations of the first machine. On the other hand, a job should start its execution on the second machine at its completion time on the first machine, and vice versa, this scenario enforces a no-wait constraint between the two machines. Our objective is to achieve the lowest possible total completion time, i.e. minimizing the makespan.
We initially prove that this scheduling problem is NP-hard in the strong sense, indicating that finding an optimal solution is computationally demanding. However, we identify some specific cases that can be solved in a polynomial time. For the general problem, we introduce three lower bounds to estimate the minimum makespan and propose two primary resolution methods: a new heuristic approach (H1) and a genetic algorithm (GA). Additionally, two different hybridizations are proposed between H1 and GA. Computational experiments have been conducted in order to assess the performance of the proposed algorithms.