5

enter image description here

the code s from a recommendation engine using surprise module, i can't find the answer anywhere.

Raj Singh
  • 408
  • 1
  • 7
  • 17
  • This link has another way of doing it ( by using KFold) https://stackoverflow.com/questions/61305393/svd-has-no-split-attribute – sakeesh Sep 08 '21 at 01:21

1 Answers1

4

Depending on what you objective is, you can use the cross_validation method and it will perform the splits automatically for you. Example: cross_validate(algorithm_to_use, data, measures=['RMSE'], cv=5)... The cv=5 tells the cross validation iterator how to split the data.

See documentation on Cross validation.

You can also explore the model selection package for what you need.

Context is important when trying to help you find a solution.

Serg
  • 2,346
  • 3
  • 29
  • 38
kevin james
  • 51
  • 1
  • any idea about how can i specify algorithm_to_use. [this](https://surprise.readthedocs.io/en/stable/algobase.html#surprise.prediction_algorithms.algo_base.AlgoBase) link does not help much. Better if you have link with example – Vishvajeet Ramanuj Mar 16 '21 at 05:46
  • This link has another way of doing it by using KFold [link](https://stackoverflow.com/questions/61305393/svd-has-no-split-attribute) – sakeesh Sep 08 '21 at 01:22