Recognition of a Life Distribution Based on a Neural Network

In general, we describe three different methods to select an appropriate distribution form: histogram, probability plots, and hypothesis test. The life distribution is recognized by a neural network method. The relationship among life distribution with life data is described through threshold and weight of neural networks. The method is convenient to use. An example is presented to validate this method, and the results are satisfactory.