https://github.com/nickenshidqia/predict_gpa_from_sat_score_using_linear_regression
Build a machine learning model that can predict GPA from SAT Scores to evaluate the potential academic success of applicants.
https://github.com/nickenshidqia/predict_gpa_from_sat_score_using_linear_regression
data-science gpa linear-regression logistic-regression machine-learning python
Last synced: 3 months ago
JSON representation
Build a machine learning model that can predict GPA from SAT Scores to evaluate the potential academic success of applicants.
- Host: GitHub
- URL: https://github.com/nickenshidqia/predict_gpa_from_sat_score_using_linear_regression
- Owner: nickenshidqia
- Created: 2023-12-18T12:16:23.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-12-18T14:14:20.000Z (over 2 years ago)
- Last Synced: 2025-04-10T00:46:29.775Z (over 1 year ago)
- Topics: data-science, gpa, linear-regression, logistic-regression, machine-learning, python
- Language: Jupyter Notebook
- Homepage:
- Size: 1.23 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Machine Learning Project Using Linear & Logistic Regression to Predict GPA from SAT Scores
## Project Description
**Problem :**
Understanding the relationship between standardized test scores and academic performance is essential for educational institutions to make informed admission decisions. By leveraging historical data, the goal is to create a tool that assists admission offices in evaluating the potential academic success of applicants and provides valuable insights into the predictive power of SAT scores.
**Challenges :**
Build a machine learning model that can predict GPA from SAT Scores
## Project Goal
This project aims to develop a predictive model that can estimate a student's GPA based on their SAT scores.
## Tools & Library Used
[
](https://www.python.org/)
[
](https://jupyter.org/)
## Project Result
[Click here to get full code](https://github.com/nickenshidqia/Predict_GPA_From_SAT_Score_Using_Linear_Regression/blob/74f423e3073470eee6465db06c9d6559230dd5b1/GPA%20%26%20SAT.ipynb)
### Dataset
- There are 84 students who have studied in college
- SAT Score = Critical reading + Mathematics + Writing
- GPA = Grade Point Average (at graduation from university)
### Linear Regression
#### GPA based on SAT Score
- That is the best fitting line, or the line which is closest to all observation simultaneously
- Example if there is student who has SAT score 1700, then he will got GPA 3.165
- There is strong relationship between SAT and GPA
- The higher the SAT of a student, the higher their GPA
#### GPA based on SAT & Attendance
- From this dataset, we found that average of students attendance more than 75% of lectures is only 46.42% have attended. Mean < 0.5 shows that there are more 0s than 1s.
- On average the GPA of those who attendeded is higher than the one didn't attend the class.
#### Making Predictions
**Prediction 1**
Create prediction of 2 students, whose the one that get higher GPA :
- Budi, who got 1700 on SAT and did not attend >75% of lecturers
- Ani, who got 1670 on SAT and attended >75% of lecturers
- The predicted GPA at graduation for Budi is 3.02
- The predicted GPA at graduation for Ani is 3.20
- Ani scored lower on SAT, but she attended > 75% of lectures, and she is predicted to graduate with a significantly higher GPA than Budi.
**Prediction 2**
Create prediction of GPA for SAT score 1740 and 1760 :
- The predicted GPA for SAT score 1740 = 3.155938
- The predicted GPA for SAT score 1760 = 3.189051
- The higher SAT score, the higher GPA score
### Logistic Regression
#### Predicting whether student will be admitted or not
- This function shows the probability of admission given an SAT score
- When SAT score is relatively low, the probability of getting admitted is 0%
- When SAT score is relatively high, the probability of getting admitted is 100%
- Score between 1,600 and 1,750 is uncertain
- SAT score 1,650, the students roughly 50% chance of getting in
- SAT score 1,700, the students got 80% chance of getting in
#### Predicting which gender will be the most admitted
- odds of female to get admitted are 6.99 times odds of male
- given the same SAT score, a female has 7 times higher odds to get admitted than the male
- in this particular university (degree), it is much easier for females to enter
- example communications, most of them are female, while STEM predominantly male
### Accuracy
- The accuracy of our model is 94.64%. Our model seems good at classifying