Neural network data modeling and its application on effect analysis of head and shoulder acupoint massage

In this paper, a neural network (NN) data modeling method is proposed to analyze the relationship between the inputs and output of a complex system. Meanwhile, the data analysis method is applied on the scientific study of head and shoulder meridian acupoint massage. First, the NN data model is introduced by involving relationship weightings and trained by the least-square method. Next, the experiments of head and shoulder massage are performed by massaging five important meridian acupuncture points in a standard manner, while the heart rate variability of each subject is evaluated before and after the massage. Then, the NN data analysis indicates the validity and effect of the massage, and the result is checked with a statistical method. Thus, the stress relaxing effect of massaging head and shoulder meridian acupoints can be obtained from the NN data analysis.