- Load the testing data (i.e. testing_set_label.csv) and name it as
new_test_data. You can load the data using the below command.
new_test_data = pd.read_csv("testing_set_label.csv" )
Here the target column is deliberately not there as you need to predict it.
- With the model you have built, predict the target variable on the testing set that you loaded in step 1. Store the predicted values of the testing set in a variable target.
- Save prediction result using the guidelines which are given below this section “how to save prediction file?” section.
- Submission: Upload the prediction file in CSV format under “submit” tab on the challenge page. Note: Make sure that the csv has the same header names as you are given in the test data.
- Once done, click on the ‘Make Submission’ button. You’ll be able to see your submission with the evaluation score.
Please follow the steps given here: Datathons/Saving_prediction_file.ipynb at master · dphi-official/Datathons · GitHub
Once you save your predictions in csv, you can simply upload it under the “submit” tab of the problem and check your score against an evaluation metric.
There is no limit on the number of submissions. One can do any number of submissions each day until the deadline of the Data Sprint.