You are a data engineer working on optimizing a Spark application to improve its performance on a large dataset. The application is experiencing significant delays, and you suspect that data shuffling is a major bottleneck. The Spark UI provides various tabs that offer insights into different aspects of the application's performance. Considering the need to minimize costs while ensuring compliance with data processing standards, which of the following strategies would you employ to identify and mitigate performance bottlenecks related to data shuffling? Choose the best option from the four provided.