A platform that integrates a predictive model using machine learning to flag students at risk of dropping out of high school. Based on significant evidence supported by research, the model uses isolated inputs recognized as predictive markers for students dropping out and thus assigning a percentage of probability at which they are likely to drop out. More importantly, the platform provides a user-friendly interface that allows for the data to be easily visualized allowing for the school to identify students at risk of dropping out. The feature enables schools and parents to intervene prematurely to prevent such actions to occur.


We recognized a need to reduce drop out rates in high schools as there was a recognizable correlation between the dropouts and there negative impact on society and their respective communities.

What it does:

An early intervention platform for potential high school dropouts.

How we built it:

Using machine learning that incorporates the use of random forest algorithm coded in Python, as well as JavaScript for website building, and protocol design including API requests.

Challenges we ran into:

The main challenge we faced was the absence of a data set that is required to be run through our predictive model. Another issue was protocol design, specifically getting the backend and frontend to communicate.

Accomplishments that we're proud of:

To get around the obstacle of restricted data, we generated our own dataset to be used as a placeholder. Moreover, creating a web application with very limited pre-existing knowledge of JavaScript and it with our backend with very limited pre-existing knowledge in protocol design.

What we learned:

Understanding random forest decision trees within Machine Learning, python libraries, GitHub repositories, Google sheets API, protocol design, JSON object, HTML, and JavaScript.

What's next:

There a couple of things we hope to work on! For one; retrieving the proper dataset by partnering with schools, adding messaging hubs, a better user interface and data visualization on our platform.

Built with:

Python, JavaScript, HTML/CSS, Google Sheets and Forms, JSON, GitHub, git, and Machine Learning; Random Forest, Decision Trees, Lasso, and Extra Tree Class.

Prizes we're going for:

Intel® Movidius™ Neural Compute Stick

$100 Amazon Gift Cards

Hustle Award

Social Entrepreneurship Award

Grand Prize

Jetbrains Pro Software

Blu R2 Plus Smartphones

Misfit Shine 2

Team Members

Amartya Singh, Jaskaran Singh, Robert Leifke, Xuan Chau
View on Github