Back To Projects

Texas High School Dropout Rates

Emily J.

Dropout rates in high schools throughout Texas have been steadily increasing, especially following the Covid-19 pandemic. This project aims to solve this problem through an artificial intelligence algorithm that can predict the dropout rate of a Texas high school campus based on specific characteristics of the school.


Dropout rates in high schools throughout Texas have been steadily increasing, especially following the Covid-19 pandemic. Dropping out of high school has a permanent effect on the future of a person’s life, and as more students don’t make it to graduation, the more serious this issue becomes. The only way to combat this problem is by identifying what factors or characteristics of a campus are causing students to drop out. This project aims to solve this problem through an artificial intelligence algorithm that can predict the dropout rate of a Texas high school campus based on specific characteristics of the school. This project uses a gradient boosting model on a dataset by the TEA (Texas Education Agency) with information about high schools throughout the state. The results of this model are also implemented into an app where people can input information about a school and get a dropout rate prediction, as well as recommendation for that specific campus on how to decrease dropout rates.

Explore More!

Source Code
Project Webpage
Emily J.
Philip Bell
Computer Science MS from Georgia Tech, Harvard and Oxford alum, AI Engineer at Cognino

Related Projects

Standardized Testing is it Really Effective?

How effective is standardized testing in assessing students' knowledge? With data from previous students on sat scores and GPAs from college and high school, I can use this data and plug it into different models to determine how well these sat scores determine GPA.
Rohan S.
Mentored by Christina Cheng
Detecting Facial Expressions

This paper shows that machine learning algorithms can be used in order to classify emotions into different categories. The most promising results were from the K-nearest neighbor model, with 61.5% accuracy in the training set, and 32.1% accuracy in the testing set.
Bertran M.
Mentored by Odysseas Drosis
A new style of teaching: Exploring the benefits of visual language learning

This project was designed to investigate the efficiency and likelihood of people learning a new language through means of an AI that would help translate objects in day to day life to a language that the user would want to learn.
Yuna S.
Mentored by Shreya Parchure