Hi, everyone I am working on a regression problem Why my model accuracy is showing zero for tabular data with tabular learner
I’ve only just started deep learning so I could be wrong, but I believe accuracy measures how many predictions the learner gets exactly right. If the dependent variable (y) is continuous and there are many variables that affect it (x), the chance of getting a prediction 100% right would be very low and therefore accuracy would most likely be 0. You would be better off looking at MSE, RMSE, and R2.
1 Like