Hi, your video is very informative and full of explanation. However, as I know that the rule of thumb is: never mess up with your test set. Always split into test and train sets BEFORE trying oversampling/undersampling techniques! Because oversampling before splitting the data can allow the exact same observations to be present in both the test and train sets. This can allow model to simply memorize specific data points and cause overfitting and poor generalization to the test data. Data leakage can cause you to create overly optimistic if not completely invalid predictive models. So what you think of that?
Hello, how about if we have a panel data of clients over time? And may be we want to see the impact of changing salary on the decision to stay or quit the bank? Can we do it, please?
First of thanks for your comment. Dear, we have trained our model on available features just for "future predictions."(just for training our model) by removing customer ID, which is not affecting our target variable. Suppose a new customer comes. Because we know who he is. We have to input different features to our model for prediction by removing customer ID. We can check the churn rate for this particular customer. hope I'm clear. Let me know if you still have some queries.
Check my code here : github.com/DataThinkers/Machine-Learning-Projects-Code/blob/main/Predicting%20Employee%20Churn%20Using%20Machine%20Learning%20(1).ipynb
@@DataThinkers, I opened this link. but no any ipynb, CSV, Excel file is available. when i click on Churn_Modelling.csv looks like code. Please check once. this project helps me in my interview. please help me🙏
Churn_Modelling.csv is a dataset file (in CSV file format) just download it and use it Here is the CSV file (Dataset) : github.com/PRIYANG-BHATT/Datasets-RU-vid-Pandas/blob/main/DS/Churn_Modelling.csv Here is the code file : github.com/PRIYANG-BHATT/RU-vid_Machine_Learning_Projects_code/blob/main/Project%20-%204%20Bank%20Customers%20Churn%20Prediction.ipynb Best of Luck
Dear Priyang, Thank you for your precious teaching . Please do provide us a video on python Tkinter for Gui. Also while working out the same i ma getting a comment as "/usr/local/lib/python3.10/dist-packages/sklearn/base.py:432: User Warning: X has feature names, but Logistic Regression was fitted without feature names warnings. warn( ", . Iam getting the same type of comment in all the models, Because of this iam getting differences all the predictions, Request your earnest support.. Thanks in advance.. God bless.
Thanks, use and run my code once, also compare it with your code:github.com/PRIYANG-BHATT/Machine-Learning-Projects-Code, sure i will upload on video on tkinter.
@@DataThinkers Thank you so much and appreciating your quick response. When i was working with your codes everything was working quite right and smooth. Thanks with gratitude for your timely support. God bless....