In machine diagnostics, it is difficult to collect for learning all possible operating modes of machine functioning. Some operating modes will usually be missing. In these circumstances, it is important to know which modes (subclasses) are the most valuable for successful machine diagnostics. It is also of interest to investigate the usefulness of noise injection to cover missing operating modes in the data. In this paper, we study the importance of selecting different operating modes of a water-pump and using them for learning in 2-class and 4-class problems. We show that the operating modes representing different running speeds are more valuable than those representing machine loads are. We also demonstrate that 2-Nearest Neighbors directed noise injection is useful when filling in missing operating modes in the data.