A total of more than 700 data for training, two classification problem, the trained model after save, with another set of similar data to predict and forecast precision is low, with keras, input data have been disrupted, training and prediction of data through the same processing, also is the issue of CNN parameters

CodePudding user response:

Data may lead to a less fitting, it is suggested that enhanced do data

CodePudding user response:

Is solved, please I have tens of thousands of data set, but the accuracy of the prediction is also very low