Optimality Benchmark for Combinatorial Optimization

This dataset is an optimality benchmark for 1 synthetic 3 real-world application scenarios:1. A synthetic dataset2. Residential energy consumption3. Bike sharing4. Charging control of electric vehiclesThe dataset consists of 1 million random network positioning of agents in a binary tree, which are used in the collective learning algorithm of I-EPOS to explore the learning capacity of the combinatorial landscape. This dataset can be used as reference of other heuristic algorithms and enhancements. Besides random positioning, the dataset comes with 124 metrics that evaluate deterministic criteria for the agents' positioning.