Campus Placement Analysis and Prediction

In College selection time, Every Student First Question is what is the percentage of campus Placement ?

Because Campus placement is one of most major things for Student to get a better career.

Even more than just that. This program would tell which things affect your placement? Sounds great, right? Let’s look at how we can accomplish this with the help of Data Analysis!

and predict the Percentage of Placement campus with the help of Machine Learning!

Overview of the Problem…

What is in it?

This data set consists of Placement data of students in Jain University Bangalore.

It includes secondary and higher secondary school percentage and specialization. It also includes degree specialization, type and Work experience and salary offers to the placed students

You can download the data set from kaggle datasets. and analysis the data and answer these following question:-

Business Understanding

Questions

1. Which factor influenced a candidate in getting placed?

or

Does percentage matters for one to get placed?

2. what is the percentage of female or male how get a placement?

3.Which degree specialization is much demanded by corporate?

4.Prediction model using different ML algorithms?

Analysis

Analysis the data for cleaning and finding some Questions Answer.

Data Understanding

Here are a Steps to find the Question Answer

I am using python in jupyter-notebook to analyze the data.

First we load important libraries and the student dataset.

This data set consists of Placement data of students in campus. It includes secondary and higher secondary school percentage and specialization. It also includes degree specialization, type and Work experience and salary offers to the placed students

This data set consists columns are
 sl_no          : serial number
 gender         : [male,female]
 ssc_p          : means--> 10th class percentage   
 ssc_b          : which board to passed out 10th.    
 hsc_p          : means--> 12th class percentage    
 hsc_b          : which board to passed out 12th.    
 hsc_s          : which stream he choose (science,commerce,arts) 
 degree_p       : means--> Bachelor degree percentage
 degree_t       : which stream choose for bachelor 
 workex         : It has a work experience or not.
 etest_p        : entrance test percentage
 specialisation : Master degree in Mkt&HR or Mkt&Fin 
 mba_p          : Master degree percentage
 status         : He/She got placed or not in campus placement.    
 salary         : placement packages.
sscb- Secondary School Certificate Board-> Which is lower secondary school(1st-10th Grade)
hscb- Higher Secondary Certificate Board-> Which is higher secondary school(11th and 12th Grade)
Grade represents their year of study in school.
Boards are different curriculum/education system adapted by educational institutions.

First of all, we know the role of the campus placement in students life.So with the help of this student placement data, find some important factor which is important in placement.

Some insights from data exploration are:

No of students placed or not placed

See the above Graph the number of placed student is 148 and not placed student is 67.But we not know how much male or female in this graph so we see another graph that show exact no of male or female

2. what is the percentage of female or male how get a placement?

No. of female or male who placed or not.

And also find the percentage of placed male or female. and not placed student percentage.

Now check the correlation between diffrent feature with placement status and find…

1. Which factor influenced a candidate in getting placed?

output:-

ssc_p to placement  60.8 %
hsc_p to placement  49.1 %
mba_p to placement  7.7 %
degree_p to placement  48.0 %
etest_p to placement  12.8 %
workexp to placement  27.6 %

hence we see that the academy percentage most important role to get a placement .

2. Does percentage matters for one to get placed?

Show the below graph and answer

We can see that those placed have slightly higher degrees that those not placed. Therefore, grades could be said to be important.

3. Which degree specialization is much demanded by corporate?

we see the bar graph and easily to say that which specialization in demanded by corporate.the marketing and finance specialization is more demanded then marketing and HR .

Modeling

4. Create model Using Machine Learning

To find better model experiment with different ML method to best result.

So I am experiment with Logistic regression and Random Forest Classification. then the accuracy rate is

In Logistic regression method find accuracy

Using Logistic Regression Method
Accuracy 90.70%
              precision    recall  f1-score   support

           0       0.89      0.89      0.89        18
           1       0.92      0.92      0.92        25

    accuracy                           0.91        43
   macro avg       0.90      0.90      0.90        43
weighted avg       0.91      0.91      0.91        43

In Random Forest Classification method find accuracy

using Random Forest Classification method
Accuracy 86.05%
              precision    recall  f1-score   support

           0       0.83      0.83      0.83        18
           1       0.88      0.88      0.88        25

    accuracy                           0.86        43
   macro avg       0.86      0.86      0.86        43
weighted avg       0.86      0.86      0.86        43

after using different ML algorithm we predict

In Logistic Regression Method find the accuracy of prediction is 90.70%. And the random forest classification the accuracy of prediction is 86.05%.

Thus we can say that logistic regression model is better perform then Random Forest Classification model.

Finally, Now it’s Your time try to analysis and visualization your own collage placement datasets, and show which is important feature in placement.

I am very grateful for my first opportunity to contribute something to the great Data Science community. I am looking forward to sharing more of my work and connect with all of you.

If you want to see the full code, here is the GitHub repository.