A Dataset for Multi-Server Multi-User Computation Offloading

In the instruction pdf, we give the dataset format, as well as the relevant feature numerical ranges and statistical analysis of the dataset files. For each scale setting, there are two files with the same name in the dataset. The yaml file is the global parameter setting, and the txt file is the data body. Due to the large scale of the problem and the high complexity of the solution, all data files without the prefix "refine_" are suboptimal datasets. The difference in global cost after optimization with the optimal result is mostly less than 10%. The data files with the prefix "refine_" are small-scale truth value datasets of the corresponding scale data files. Each line in txt is a sample. The attribute names introduced in the dataset format are prefixed with "gt_" for labels, and the others are features. Therefore, when using it, just read the features as a directed graph with node features and edge features. The labels and features correspond to the same graph, but only some edges are retained in the labels as the result of offloading decisions and computational resource allocation. If necessary, you can also use the scripts we provide to read and preprocess the data.