Credit Risk - Probability of Default, End-to-End Model Development | Beginner to Pro Level

Credit Risk Modelling | End - to - End Development of Probability of Default Credit Risk| Kaggle Competition Data
Banks play a crucial role in market economies. They decide who can get finance and on what terms and can make or break investment decisions. For markets and society to function, individuals and companies need access to credit.
Credit scoring algorithms, which make a guess at the probability of default, are the method banks use to determine whether or not a loan should be granted.
In this video we have started such development of PD model, absolutely from a beginner level and have explored the data widely to come up with a conclusive data which fits best for the model and have developed the model using XGBoost.
It is expected that you have some basic understanding of Python and Banking domain and after watching this video you would be in very good position to pick any given data and start with your own development.
You might also like to watch -
Exploratory Data Analysis - • EDA - Exploratory Data...
Credit Card Fraud Detection ML Model - • Data Science | Credit ...
Logistic Regression Model Development - • Mastering Logistic Reg...
Linear Regression Model Development - • Data Science | How to ...
Decision Tree Model Development - • Machine Learning Proje...
Data Science Full Playlist - • Data Science
URL for Kaggle competition discussed in the video - www.kaggle.com/c/GiveMeSomeCr...
To Download the script created in the video -
File Name - PD_credit_risk_model_dev.py
URL - github.com/LEARNEREA/Data_Sci...
credit risk,
credit risk modelling,
credit risk analysis,
Banking, XGBoost
IFRS9, Basel
#CreditRisk #ProbabilityOfDefault #Banking #MachineLearning #PythonProgramming #XGBoost

[@K-mk6pc]

Trained on the same data and predicted on the same Data

[@nikhilbothra3027]

model building starts at 58:10

[@AmitSharma-sc9gg]

Many thanks for the video.
@32.25
RevolvingUtilizationOfUnsecuredLines
This should be between 0-1, why can't we simply replace by 1 all the values >1 ?
my apologies if u have explained this further in the video , just thought of asking while watching the video.
Thanks again

[@asharma1775]

Thanks . It's great

[@learnerea]

Glad you liked it

[@Camila-fv9qj]

Can you explain or solve this competition in kaggle called ( Data Science for Good: City of Los Angeles ) just EDA part i saw the solution but the converting part from txt to csv it's difficult for me especially converting 60 txt files to csv

[@Mr.Arshad-oc1fd]

Thanks 👍

[@learnerea]

Thank you too

[@kanishktolani1177]

How did you come up with the intuition that SeriousDlqin2yrs and MonthlyIncome are same for some records?

[@21121990jay]

Thanks for the video , just curious about customer demographic information.Shouldn't we ignore it while building model as it fall under fair and unbiased lending practices!?

[@learnerea]

You are right, for sure we should, if the police prohibits that.

[@amirthalakshmik453]

For this model , how can we give a manual input from user to check the output respective for it rather than the training and testing data.

[@learnerea]

In any case first you would require to train your model so that it could understand what exactly needs to be achieved, once trained, you would for sure like to test whether the training is correct or not.
Yes, after this training and testing work for sure you can integrate the model with some application development where user can enter the value and the model give output.
for the purpose of practicing, you can use Input function in python to prompt for entering and input and basis on that it produces the output, a demo -
ua-cam.com/video/XS1fR3WJj4I/v-deo.htmlsi=WB37bIkU5vVBW_cf

[@biplabkar1391]

Where can I get the data set?

[@learnerea]

from the given Kagle URL in the description

[@tanumoyhazra6055]

Very informative video.can i get the python codes ?

[@learnerea]

Link in the description

[@K-mk6pc]

It seems the Github repo has been moved, We cant find the Source!!!!

[@learnerea]

Will check out and upload again

[@Anonymous_Pers0n]

so dry and no visualizations.

[@editingontheway7318]

the video was great but i think u messed a bit in the last by thinking u made a long video
but thank you i watched the whole video as u said i got a general idea and the way u approached using those methods helped me a lot🙌

[@learnerea]

I'm glad it was helpful