0% found this document useful (0 votes)
30 views11 pages

Student Dropout Prediction Project

The project involves analyzing a dataset of 5,000 JEE aspirants to predict student dropout rates after Class 12 using exploratory data analysis (EDA) and K-Nearest Neighbors (KNN) classification. Students will utilize tools like Google Sheets or Excel to clean data, perform EDA, and implement KNN and K-Means clustering. Deliverables include answering assessment questions and submitting a comprehensive report detailing their analysis and findings.

Uploaded by

bbvyas4udemy
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
30 views11 pages

Student Dropout Prediction Project

The project involves analyzing a dataset of 5,000 JEE aspirants to predict student dropout rates after Class 12 using exploratory data analysis (EDA) and K-Nearest Neighbors (KNN) classification. Students will utilize tools like Google Sheets or Excel to clean data, perform EDA, and implement KNN and K-Means clustering. Deliverables include answering assessment questions and submitting a comprehensive report detailing their analysis and findings.

Uploaded by

bbvyas4udemy
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Week-1: [Link]

feature=share

Week-2: [Link]

Week-3: [Link]

Week-4: [Link]

Week-5: [Link]

Week-6: [Link]

Week-7: [Link]

Week-8:
Project Overview

In this project, you will work with a student academic and behavioural dataset to predict whether a student is likely to drop out after Class 12. The goal is to
apply exploratory data analysis (EDA) to gain insights from features and implement the K-Nearest Neighbors (KNN) algorithm for classification. This project
can be completed using Google Sheets or Excel, though students are free to use other tools if they prefer.

Project Requirements

Tools

 Google Sheets (or Excel)

 Any other tool (optional)

Skills Covered

 Data Cleaning

 Exploratory Data Analysis (EDA)

 KNN Classification

 K-Means Clustering
Project Deliverables

You will be required to answer the assessment questions available on the portal and also submit a comprehensive project report detailing your approach,
analysis, and findings.

Students are encouraged to be creative in their reports while ensuring all the required sections are covered.

Dataset

Link to the dataset: Refer attached spreadsheet

The dataset contains information about 5,000 JEE aspirants with the following columns:

 jee_main_score: Score obtained in the JEE Main exam (0–100 scale).

 jee_advanced_score: Score obtained in the JEE Advanced exam (0–100 scale).

 mock_test_score_avg: Average score across mock tests taken by the student (0–100 scale).

 school_board: Educational board of the student (CBSE, ICSE, or Other).

 class_12_percent: Percentage score in Class 12 board exams.

 attempt_count: Number of attempts the student has made for JEE (typically 1 or 2).

 coaching_institute: Type of coaching attended (FIITJEE, Local, or Other).

 daily_study_hours: Average number of hours the student studies per day (0–10 scale).

 family_income: Economic background of the student (Low, Mid, or High income).

 parent_education: Highest educational qualification among parents (Upto 10th, PG, or Other).

 dropout: Target variable indicating if the student dropped out after Class 12 (1 = Dropped out, 0 = Continued).
As a data scientist, you are tasked with analyzing this dataset. Your objective is to:

 Conduct an exploratory data analysis (EDA),

 Implement K-Nearest Neighbors (KNN) classification for prediction, and

 Apply K-Means clustering to group students.

Accordingly, this project's problem statement is divided into three sections. Please answer each question one by one.

Section 1: EDA

[Link] many students have 100 in class_12_percent?

a. 20

b. 22

c. 24

d. 26

2. What is the average of class_12_percent? (Mark the closest value)

a. 75

b. 50

c. 60
d. 70

3. What is the most frequent value in location_type?

a. Urban (1652)

b. Rural (1671)

c. Semiurban (1680)

d. None of these

4. What is the median value of daily_study_hours? (Mark the closest value – 4.48386)

a. 1

b. 2.5

c.3

d. 4.5

5. How many students have mental health issues?

a. 2358

b. 2574

c.2426
d. 2635

6. How many students whose location_type is Urban and they dropout?

a. 123

b. 246

c.369

d. 1282

7. How many students attempted more than once?

a. 2523

b. 2477

c.3422

d.2387

8. Which school board has the highest average JEE Main score?

a. CBSE – 71.96082278

b. ICSE – 71.73312651

c. State – 72.19802122

d. Can not be determined


MOST IMPORTANT
Section 2: K-NN

Consider a new student record with the following details for dropout prediction:

jee_ma jee_adva mock_test school class_1 attemp coaching daily_st family parent_ locati peer_pre mental_h
id in_scor nced_sco _score_av _boar 2_perce t_coun _institut udy_hou _inco educati on_ty ssure_lev ealth_issu
e re g d nt t e rs me on pe el es

ST1
000 61.26 65.14 71.535 ICSE 71.85 1 Local 4.05 Mid PG Rural High Yes
1

You want to predict the dropout based on the data.

Note:

Before calculating the Euclidean distance, convert the categorical columns into numerical values as follows:

 school_board: CBSE = 0, ICSE = 1, State = 2

 coaching_institute: No Coaching = 0, Allen = 1, FIITJEE = 2, Local = 3

 family_income: Low = 0, Mid = 1, High = 2

 parent_education: Upto 10th = 0, 12th = 1, Graduate = 2, PG = 3

 location_type: Rural = 0, Semi-Urban = 1, Urban = 2

 peer_pressure_level: Low = 0, Medium = 1, High = 2

 mental_health_issues: No = 0, Yes = 1
9. Using the encoded dataset, compute the Euclidean distances between the new student (ID = 10001) and all existing students' Data. Based on these
distances, which 3 nearest neighbors (Student_IDs) are most likely to predict the dropout of new student?

a. ST5003 - Yes

b. ST5751

c.ST6734

d. ST8900

e. ST5732 - Yes

f. ST9222

g. ST9394 - No

10. Using K-NN (K = 5), predict the dropout for the new student (ID = ST10001). Enter 1 for Yes and 0 for No

11. After encoding categorical variables, apply K-NN algorithm with (K = 20). How many of the 20 nearest neighbors have a dropout = 0.?
Section 3: K-Means Clustering

You are applying K-Means clustering with ( K = 2 ) to group the student records. You initialize the clusters with the applicant having
ID "ST7612" and "ST9269".

Cluster center id

C1 ST7612

C2 ST9269

12. Consider the student record having ID: ST5001. What is the distance between this student and cluster center C2 (ID: ST9269)?

Note: You have to ignore the id and dropout columns. Use encoded dataset to calculate the Euclidean distance.

a. 44.64

b. 50.63

c. 30.44

d. 80.21

e. None of these

13. Which cluster would the record having ID : ST5745" be assigned to after the assignment step?

a. Cluster 1 (7612)

b. Cluster 2 (9269)
14. After the assignment step, perform the cluster recomputation step to obtain the new cluster centres. What are the new values of C1 and C2?

a.

mock_test school class_12 attemp coaching


jee_mai jee_advan daily_stu family_ parent_e locatio peer_pres mental_he
_score_av _boar _percen t_coun _institut
n_score ced_score dy_hours income ducation n_type sure_level alth_issues
g d t t e

80.52 86.22 81.77 1 65.31 2 3 3.4 2 0 2 1 1

90.14 73.48 54.98 2 70.62 2 3 6.7 1 3 0 1 0

b.

mock_test school class_12 attemp coaching


jee_mai jee_advan daily_stu family_ parent_e locatio peer_pres mental_he
_score_av _boar _percen t_coun _institut
n_score ced_score dy_hours income ducation n_type sure_level alth_issues
g d t t e

50.29 72.98 71.89 2 78.33 1 1 5.5 1 3 2 0 1

88.63 68.29 74.08 1 73.05 2 3 5.9 2 1 0 1 0

c.
mock_test school class_12 attemp coaching
jee_mai jee_advan daily_stu family_ parent_e locatio peer_pres mental_he
_score_av _boar _percen t_coun _institut
n_score ced_score dy_hours income ducation n_type sure_level alth_issues
g d t t e

97.37 40.03 92.08 2 79.76 1 2 4.6 2 0 2 2 0

58.99 67.25 94.43 1 76.58 2 2 1.4 1 3 2 2 0

d.

mock_test school class_12 attemp coaching


jee_mai jee_advan daily_stu family_ parent_e locatio peer_pres mental_he
_score_av _boar _percen t_coun _institut
n_score ced_score dy_hours income ducation n_type sure_level alth_issues
g d t t e

71.82 40.53 70.85 0.95 75.73 1.49 1.52 4.55 0.97 1.47 0.99 1.02 0.47

71.98 69.87 67.66 0.98 74.83 1.51 1.52 4.47 1.00 1.49 1.00 1.00 0.49

You might also like