Train and Test Dataset in ANN workflow

I have a dataset of training with three columns and last columns with rating. Now for the test my rating column is empty. Because I want to predict the rating in test set. In partitioning the data, I use “Take from top”. As I want to sent the empty column of “rating” for prediction. When I train and test this dataset, I get very low accuracy like 54%. How can I improve accuracy?

Hey there,

Mind sharing your workflow and sample data? Without more specifics it is really difficult to help you.

2 Likes