Project77 partnered with a Higher Education Graduate Program to help distribute its students into diverse class clusters and teams by creating a functionalized matching algorithm and data reporting dashboard. The current state of creating clusters for the higher education stakeholders required the manual “drafting” of 1000+ students into clusters. In creating diverse clusters, the individuals had to weight multiple variables including (but not limited to) race, nationality, gender, quantitative skillset, and previous work experience. The process was performed by 3 individuals over the course of 200+ hours. Moreover, there was a 28% error rate based on the matching requirements determined by the University, which would come to light the following year as the stakeholders had no capacity to evaluate the effectiveness of the clustering. Project77 created a matching algorithm to form unique groups. Using a greedy approach, each cluster (group of students from the entire class) and learning team (smaller groups within a cluster) was optimized for diversity across predetermined criteria. At a glance, imagine every student is a marble. The total class is a bag of marbles. Some marbles appear to look exactly the same. Each step of the way a marble is selected and compared to each cluster or learning team, and if the cluster or learning team will benefit, then that student will be placed in that group. If the student is not selected, then he or she will be put back into the marble bag until another set of criteria is compared. Algorithm 1. Created unique profiles for students based on matching requirements criteria Results Faster: Created clusters in seconds Quality: 3% error rate (this was optimal based on data set) Functionalized: Dynamic (easy to change weights and swap variables) Project77 empowers organizations with data insights and services to help drive decisions. Founded by social entrepreneurs and leaders in the education space, we bring cross-sector solutions to our partners through our expertise in using data analytics to drive best practice. Project77 utilizes a variety of data techniques to glean meaningful insights from data. We specialize in statistical techniques such as linear/logistic regression, time series analysis, and unsupervised learning models (e.g. clustering, hierarchical clustering, k-means) and prescriptive analytics techniques such as Difference-in-Differences, Monte Carlo simulations, and optimization (i.e. Solver). For big data to be useful, the refined information needs to be able to bridge the gap between data silos and decision makers. Project77 works very closely with its partners to create customized reporting that end users can interpret and act upon. Additionally, we ensure that any outputs created can easily be integrated with current processes. Project77 utilizes a variety of software packages to create visualizations that take complex findings and present them in a way that is informative and engaging. As the size and scope of data increases, the complexity of the data analysis increases as well. Project77 synthesizes the analysis and insights of data with powerful data visualization packages to produce valuable, meaningful, and actionable outputs. Project77’s knowledge of popular programming languages used in data analytics allows us to tackle data issues with unprecedented approaches towards statistical analysis, automation, and efficiency development. Project77 strongly believes in a hands-on approach when cleaning, transforming, and interpreting data. We want to ensure a deeper and more insightful understanding of the data to encourage a well-informed decision-making process. Our skillsets not only allow us to process complex educational data, but also to effectively communicate information back to all relevant stakeholders. The common thread that links high performing organizations is the ability to build strategic processes, measure effectiveness, learn insights, and continuously adapt to better serve stakeholders. We strive to help make education organizations agile, by identifying problem areas and barriers to student success, creating sustainable measurement frameworks based on data, and providing targeted recommendations. Our aim is to work with our partners to provide long-term solutions that dramatically impact student learning and success. Higher Ed Student Experience
OVERVIEW
About Project77
Data Aggregation, Feature Selection, and Model Selection
System Implementation and Customizations
Intuitive Data Visualizations, Frameworks, and Reports
Programming Languages Used: Python, SQL, R, and VBA
Personal Data Coaching, Change Management, and Technical Support
Why We Do What We Do
Project77 partnered with a Higher Education Graduate Program to help distribute its students into diverse class clusters and teams by creating a functionalized matching algorithm and data reporting dashboard. Project77 empowers organizations with data insights and services to help drive decisions. Founded by social entrepreneurs and leaders in the education space, we bring cross-sector solutions to our partners through our expertise in using data analytics to drive best practice.
Higher Ed Student Experience
OVERVIEW
About Project77
