Try out our website here: https://maruthivemula.github.io/OptiFind/
Inspiration
As US high school students, we are committed to improving the education systems within our country. We want to make sure students from all across the nation have access to educational resources that enable them to develop into the leaders and innovators of the future. Due to real-world funding limitations for school programs and resources, it is imperative for educational institutions to better understand how they can effectively distribute funding for various educational categories which include but are not limited to transportation, community service programs, tech supplies, instructional support, academic programs, general and school administration, instructional staff, food services, and more. Through OptiFind, we are working towards providing insights for school counties to provide high quality education for students across the nation.
The market for school funding solutions is still relatively new, but it is growing rapidly. According to a report by the Education Data Initiative, public K-12 education expenditures total $667 billion. Funding solutions are a significant part of that market. OptiFunds aims to be a leader in this space, with a growing market share. We hope to create a strong reputation for providing effective data-driven solutions to schools.
OptiFunds operates in a growing market for school funding solutions with several competitors, which include EdBuild, EdOps, and SchoolSmart. However, OptiFunds stands out because rather than proposing how to redistribute funding across counties, which is already an explored problem, our product focuses on creating an expenditure plan for various academic categories based on county specific data. Our user-friendly platform and focus on this specific subproblem give us a competitive edge in the market.
Our mission is to empower schools with smart funding solutions. This is done through optimizing the use of financial resources and improving student educational outcomes. As students ourselves, we believe that every student in the country deserves access to a high-quality education. Effective financial management and expenditure distribution is essential to achieving this goal. Our platform is designed to help schools navigate through performing data-driven financial decisions, specifically through reducing waste and inefficiency by allocating resources more effectively. We are firmly committed to making positive impacts on our education system and the lives of students through our platform.
What it does
The main component of our solution is a data dashboard website that users can easily use to generate county-specific education spending reports. Once they access the URL here, , they are taken to a landing page. They can then click the “Try OptiFind Out!” button to be taken down to the interactive dashboard. The user then can look at the various counties across the nation on the map to the left hand side of the dashboard. Once they find a county of interest, they can type it into the county filter on top. Then, the dashboard will regenerate a specific report based on the data for that county. The report will display the following statistics and visualizations:
A grouped bar chart containing the current and predicted optimal expenditure proportions for various educational categories
County-specific enrollment statistics, district student performance, and population between 5-17 years old in poverty
Line plots showing the student performance metric change over time
Line plots for the total and in poverty population between 5-17 ages
A histogram for nationwide county student performance to compare with this county’s performance
Using these insights, policymakers, legislators, and educational system representatives can gain insights into how well the county is currently and historically has been performing. Additionally, they can get recommendations on how to better reallocate the funding for the county to meet the predicted optimal conditions for optimizing student performance. For example, they may identify that a county is currently spending x% of its budget on tech services but may want to increase that to y% to improve student performance. This could be in the form of providing students with computers, giving them access to educational platforms to practice concepts learned in class, etc.
How we built it
In order to get the data necessary to power our platform, we made API query requests to Education Data Explorer: https://educationdata.urban.org/data-explorer. This API pulls data from a variety of different education databases. Once we query the specific data fields we need for our dashboard from all the counties in the US, we filter out a few counties with missing data fields. After cleaning the data in Python notebooks with Pandas and NumPy, we calculated a student performance metric using graduation rate, English standardized test performances, and Mathematics standardized test performances. We calculated the optimal expenditure breakdown by looking at the 50 best performing counties in the nation according to the student performance metric. The average distribution breakdown was taken from these top-performing counties and used to determine the ideal expenditure distribution. Then, we imported the processed data, along with the optimized expenditure distribution proportions, into Tableau and created all the graphs and visualizations necessary for our dashboard. The dashboard was then exported to Tableau public and embedded onto our frontend, developed with HTML, CSS, and Javascript. To increase accessibility of our solution, we deployed it as a GitHub page, found here: https://maruthivemula.github.io/OptiFind/.
Challenges we ran into
When querying data from the Education Data Explorer API, we realized that some of the data fields were missing from some counties, especially in the states of Utah, Illinois, and Vermont. However, we did not let this limit us in our endeavors and cleaned the data to remove the counties we unfortunately could not work with. Once we did this, we performed statistical analysis on the counties and calculated the student performance metric accordingly. This was then used to analyze the top performing counties and determine the optimal expenditure breakdown.
Another challenge we ran into was because of the massive size of the big data and also missing data we were working with, we were unable to collect long-term historical trend data for all counties. Using the data we were able to pull, we made time series graphs of factors such as how the student performance metric changes over time. Studying historical data and trends is essential when determining if the county is taking the right steps towards improving its educational system and student performance.
Accomplishments that we're proud of
Although we ran into these challenges along the way, we were able to use our creative thinking skills and technical knowledge to solve them. Due to our group having a wide variety of skill sets, we were able to amalgamate our ideas and solutions to these problems to create this platform. Whether it was reading through documentation to clean our dataset or struggling to arrange the overlapping graphs on our dashboard, we were able to persevere through the challenges and come up with not just a functional solution but one we are proud of.
Through creating this project, we have not just been able to develop our technical prowess in regards to data analytics and web development but have also been able to improve our team collaboration skills in professional environments. We have been able to develop leadership skills necessary that will prepare us to make positive contributions to our world and become part of the next generation of positive change.
What we learned
As aspiring future innovators and entrepreneurs, we gained many skills throughout our experiences within this hackathon. Some of the major takeaways were learning Tableau, gaining more experience with designing aesthetically pleasing frontends, improving our data cleaning and visualization skills, and deploying web pages. However, in order to create effective solutions to the world’s most challenging problems, effective leaders must not just have the technical skills to develop solutions but also the ability to foster effective collaborative environments and have an awareness for how to identify problems. They must understand the role that our proposed solutions will have in solving them. This hackathon was the perfect opportunity to refine our skills in both of these areas. The time crunch and multitude of challenges along the way helped create a collaborative environment in which we could better understand the components of effective teamwork. Additionally, the complexity of our proposed solution allowed us to research new frameworks and platforms to create our solution on and develop technical skills we can apply in future endeavors. Overall, this was an amazing opportunity to foster the creative and entrepreneurial side within us, develop the leadership and collaboration skills necessary in changing the world, and gain the technical skills necessary to develop programs that coalesce into a functional product.
What's next for OptiFind
Now that we created our platform for OptiFind, our next step is to apply our creation to make positive changes to our world. We hope to meet with educational system representatives and legislators to inform them of how our product can be used in conjunction with policymaking and program development to improve our educational systems. Once our company grows after this initial phase of outreach, we hope to form our company into a Data Based Decision Making Consulting Company.
Additionally, we hope to apply more complex machine learning algorithms to refine our optimal distribution breakdowns. Some potential algorithms we are planning on using to extract feature importance and perform regression are neural networks, decision trees, random forest regression, support vector machines, elastic-net, and more.
Log in or sign up for Devpost to join the conversation.