Project 22 : Credit Card Fraud Detection Using Machine Learning

[@sreebvmcreation9388]

Yes , god is there, iam searching this concept, thank u bro, well done u.

[@arunchaudhary1722]

The video which deserves 221K views, ends up getting 221 views. It cannot be more saddening for a Machine learning enthusiast like me. But please carry on, hopefully some day you will become a viral youtuber (which you truly deserve). My best regards.

[@knowledgedoctor3849]

I Believe My Ganesha, All Will Be OK🌻 Thanks For Your Support, Keep Praying🌻 Har Har Mahadev 🔱

[@arunchaudhary1722]

@@knowledgedoctor3849 Har Har Mahadev 🔱

[@aniketkumre4231]

You are life saver 😊 Actually I mentioned this project in my resume and I want to clear my doubts after watching your video. I cleared all my doubts

[@knowledgedoctor3849]

I'm not bro, Mahadeva is the life saver🌻 Keep praying & support ❣️

[@aniketkumre4231]

@@knowledgedoctor3849 yes brother. Har Har Mahadev😍

[@MUKESHKUMARBCSE]

great one sir!!!

[@vinayakgarg.3769]

How to show this project as an working project?

[@felexnaddy7596]

dump model into a pickle file and use flask

[@nisargpanchal4884]

Just one doubt, you are giving all metric scores on train/test split after doing under or oversampling but in actual operation our model is given imbalenced data in real time, when i tried checking the classification report on original train/test split I was getting a bad score

[@meet3047jhgg]

here he had done oversampling before the train test split hence the data leakage had happened due to which he had got the high accuracy ,but the real flow is to oversample only on training data and then training the model and testing which will so the result that the model is good for cases of not a fraud but will behave incorrect at actual fraud cases

[@younesgasmi8518]

​@@meet3047jhggif we apply undersampling before splitting the dataset..is this correct and don't give the data leakage issue?

[@rakeshbsirvi1821]

Even I have experienced the same, I had used SMOTE to do over sampling only on training data. Results were bad when I tested various models for both validation and testing data.

[@BiswajitSibun-n4b]

Best Explanation sir

[@vinitasharma8271]

Sir i have problem with Imblearn over_sampling import SMOTE

[@ganeshthota63]

Bro can you please explain what technologies used in this project

[@tuankhainguyen5283]

Which methodology was used in this project sir ?

[@bhaveshdalvi585]

Hi, I have a doubt. Is the data cleaned or uncleaned? Also where and why do we have to make use of the class column?

[@felexnaddy7596]

it is a cleaned dataset widely available over internet. Class column is our traget column.

[@RAZZKIRAN]

thank u, great, useful

[@maximalopgaming2942]

Sir please make one new video on real time face recognition attendance system on your multiple of deep learning series please sir 🙏

[@Hgrewssauujdkhvcjjipp]

Cool 🎉

[@ritikchaurasia708]

Can you make project on the nlp

[@knowledgedoctor3849]

Yeah🌻