Mismatch in Feature Names Between Classifier's Training and Prediction Phases

[rakshit-upadhyay214]

While using the model for classification, a ValueError is raised due to a mismatch between the feature names used during the training phase (fit) and those passed during the inference phase (predict). The features passed at prediction time have inconsistent names compared to those seen during training.

Steps to Reproduce:

1. Load the trained RF_mining_model.pkl .
2. Attempt to make predictions using test data obtained from the original dataset's train-test split.
3. Observe the ValueError due to mismatched feature names.

Expected Behavior: The feature names should match and remain consistent throughout.

Proposed Solution: Training classifier model against the exact feature names as they appear in the dataset.

[github-actions[bot]]

👋 Thank you for raising an issue! We appreciate your effort in helping us improve. Our team will review it shortly. Stay tuned!

[Devanik21]

Can you specify which model is it ? GBoost or RForest

[rakshit-upadhyay214]

It was Random Forest.

[saumyacoder1709]

I want to work on this issue