A batching problem with learning effect considerations

Consider a batch-sizing problem, where all jobs are identical or similar, and a unit processing time (p = 1) is specified for each job. To minimize the total completion time of jobs, partitioning jobs into batches may be necessary. Learning effect from setup repetition makes small-sized batches; on the contrary, job's learning effect results in large-sized batches. With their collaborative influence, we develop a forward dynamic programming (DP) algorithm to determine the optimal number of batches and their optimal integer sizes. The computation effort required by this DP algorithm is a polynomial function of job size.