Fuzzy decision-making is a critical research topic in uncertain decision-making issues. Under uncertain scenarios, a group of decision makers/experts presents the fuzzy evaluation data of the criteria to an alternative. In this case, we can use a fuzzy multi-valued set (FMVS) to express them. To solve the operation problem between different fuzzy sequence lengths in FMVSs and ensure some confidence level of fuzzy assessment values from the perspective of probability, this paper first proposes a transformation technique from FMVS to a confidence neutrosophic number cubic set (CNNCS) based on confidence levels and normal distribution of fuzzy values in FMVS. Then, we present an exponential similarity measure between CNNCSs and its group DM model with some confidence levels and normal distribution in a FMVS circumstance. Finally, the developed group DM model is applied to the selection of intelligent manufacturing equipment, and then the decision results corresponding to the 90%, 95%, and 99% confidence levels reveal the decision flexibility and rationality/reliability.