i want to predict gpa of a university, which has fours semesters. hence i tried to develop four models using neural networks by coding in python language and applied keras Deep Neural Network regressor for training and testing the model. some how i was able to get mean absolute error errors in range of 0.15 -0.27 (for four models) . the accuracy of the models are in range 35%-50%. so i want to help in increasing the accuracy and decreasing the mean absolute error, also it will be greatful if i could get other model metrics( accuracy and mean absolute error ). accuracy here it measuring the predicted and actual gpa in difference of 0.1.
24 freelancere byder i gennemsnit ₹6103 på dette job
I have been working as data scientist from last 4 years while experiencing such issues everyday. Working on creating customised metric as well as sometime the given metric does not suits the problem statement.