Application on M5 Algorithm in Sensory Evaluation

With the emergence of data mining technology, as the representative of the decision tree model tree M5' algorithm has drawn increasing attention in data mining and forecasting.The learning model established by the algorithm is intuitive, clear and easy to be accepted by the user, and has a great advantage in the predicted continuous-valued. The thesis begins from the theory analysis of M5 ' algorithm, expatiates its basic principles and algorithm process. The thesis proposes a method which is using M5' algorithm to construct the learning model for sensory evaluation, and takes the practical application analysis and verification for the algorithm.