I'm trying to integrate my machine learning model (Random Forest) I did in Jupyter Notebook inside power BI. Using gridsearch , and after testing multiple algorithms, it turned out that the Random Forest was the most performant one for my dataset.
So I saved my model as a PKL file (rf_ins.PKL) and then I imported the model inside power BI using the Python visual. I used the following code:
from pycaret.regression import *
rf = load_model('C:/Users/hp/Desktop/Sprint PBI/rf_ins')
# Use the trained model to make predictions
predictions = predict_model(rf, data=dataset)
# Get the predicted label and score columns from the predictions
predicted_label = predictions['prediction_label']
predicted_score = predictions['prediction_score']
# Check if any element in predicted_label is equal to 1
if any(predicted_label == 1):
output_text = "The employee is likely to leave with a probability of {:.2f}".format(predicted_score)
else:
print(predicted_label)
output_text = "The employee is not likely to leave"
# Return the output text
output_text
I get this error:
Here's a screenshot of my data:
I'll be actually have filters (slicers) next to the visualisation , which would correspond to my features and by changing the values of the prediction , my dashboard should be as follows
I was wondering why it's not correctly working. Any help is appreciated!
