EDA - Exploratory Data Analysis | EDA on Real Life Banking Data using Python | Beginner to Pro Level

In this video we have covered -
00:00:00 - Introduction
00: 03:46 - Data overview and feature selection
00:53:41 - Feature Engineering
00:55:20 - Missing values imputation
01:26:23 - Values modification automatically in multiple columns
01:36:42 - Variables evaluation and outlier detection
01:41:15 - Binning the variables
01:57:32 - Analysing the data
02:03:25 - Univariate analysis on categorical/object/character variables
02:57:35 - Subsetting the numerical variables from a dataframe for analysis
03:03:46 - Creating correlation matrix for multiple variables | Unstacking a correlation table to make it a normal dataframe
03:16:10 - Univariate analysis on numerical variables
03:28:19 - Bivariate analysis on numeric variables
03:47:52 - Analysing the prev. application data, performing feature engineering and feature selection and merging with current app. data
04:08:54 - Doing univariate and bivariate analysis on merged dataframe
04:28:42 - Making final conclusion and preparing recommendations for bank

You might also like to watch -

Tags -
Exploratory Data Analysis,
EDA,
Feature Engineering,
Feature Selection,
Univariate Analysis,
Bivariate Analysis,
Analysis,
Loan Default Data,
Merging,
Plotting,
Seaborn Plots,
Python,
Python Programming,
Correlation Matrix

You can download the complete script and analysis points created in this video using -

#eda #PythonProgramming #EDA #ExploratoryDataAnalysis #Learnerea

[playwithvihaan]

Great work. Please make more such EDA videos with model building across industries. Thank you.

[anushkagupta]

i didn't get the steps for binning of the days employed. Can you please explain since it is difficult to understand in this video

[azadabulkalam]

Excellent job Sir. Keep going sir. In fact I have learned SAS and Python following your video. Like to recognize you as my teacher. However still I missed something and expecting that it should be added in future video. Those are like frequency table (with frequency, percentage, cumulative percentage), Cross table (with cell, row%, col%, chi square test, expected), Pivot table.

[shivammishra-hxxo]

Thank you so much to make this video. It is really a very helpful for me

[jeetkumardesai]

29:33 sir, while trying to plot, it is showing the error:
AttributeError: ‘numpy.int64’ object has no attribute ‘startswith’

How to resolve it sir ? Asap 🙏

[arjitaren]

Sir at 35:00 at correlation step there are 8 columns but correlation is of only 6 column, may be because 2 other columns have data type as object. But I am getting error but you didn't. Why?

[shubhamkurapati]

I think for cash loans the percentage is high that mean there must be more defaulters as compare to Revolving loans. please explain sir

[chayanajmera]

Hello sir. First of all excellent work you are doing out there. This project is helping me a lot in understanding EDA. At 1:54:25 you have given labels in hundred thousands but the quantile suggested all the values till 0.90 to be in the range of 10k to 70k. Can you please suggest? Thank you and stay awesome.

[Yash-wflg]

In that previous data sheet that 1 value was missing which we could ignore or impute. So, that value has Name_contract_status as approved so what can we impute on that? still median? if yes then why so? coz mean is 1 but median is 8

[krishreddy]

hi sir, i am getting 2 null values in DAYS_EMPLOYED_RANGE after binning the values, could you please help me in this

[radyoalmikyel]

what must we do for the missing value for the OBS_30_CNT_SOCIAL_CIRCLE and the other 3 variale (1021 missed) do remove or fill the by what?

[aasthagautam]

whats the need to bin the variables?
moreover, if we are binning the variables, the scatterplots are not being displayed correctly.

[kadichidu]

Please make more videos on on projects

[Yash-wflg]

what all hypothesis can be made in this dataset?

[ribhubhattacharya]

in my output the data are being truncated by "....". How to resolve that

[manoram]

35:03 I'm getting error - could not convert string to float: 'N'