Duke Big Ideas Lab Capstone Projects
To meet the energy needs of those without access, we need to know where existing infrastructure, especially transmission and distribution lines, are in relation to communities in need. Current databases track approximately 85% of global energy infrastructure capacity. The remaining 15% may dramatically impact global emissions, but are particularly hard...
This project aims to explore two topics within a world-leading surgery program. For the first project, the final goal will be to productionize a system within the existing technology stack to automate surgery scheduling. The strategy to implement this system will be to use electronic health records data to build...
The goal of this project is to develop interventions for the growing opioid crisis. To do this, the team will build a method to probabilistically fuse granular synthetic household data with publicly available data related to opioid use to predict where opioid hotspots are likely to occur, and why. The...
This project will utilize publicly available LiDAR data to estimate carbon stocks across The Conservation Fund’s nationwide portfolio of working forestlands. Students will create computational tools for processing NASA’s newest LiDAR data – Global Ecosystem Dynamics Investigation (GEDI) and utilize field data to validate and interpret data. In addition, students...
Saving Nature is a non-profit organization that works with local partners to purchase land and restore forests to connect habitat fragments in areas of high conservation concern. Part of this effort includes monitoring the area with camera traps. These camera traps can collect hundreds of images and videos each month...
The “small-watershed” ecology approach measures all precipitation chemistry (inputs) and stream solute fluxes (outputs) within a watershed. This has led to key environmental insights such as the discovery of acid rain. Currently, upwards of 150 different federally funded sites have implemented this watershed approach to evaluate site-specific questions about climate...
The goal of capstone project is to apply and extend custom analytics solutions to discover how life remains resilient in extreme environments. An explosion of data has resulted from recent discoveries of tiny single-celled life hiding out in the most extreme places on Earth. These single-celled creatures, or microbes, thrive...
Wild animals eaten by humans are known as “wild meat”, or “bushmeat” in sub-Saharan Africa. Hunting for bushmeat is both an ancient and modern practice, but as we step into the Anthropocene, bushmeat hunting has become unsustainable in many areas, threatening biodiversity and the food, financial, and cultural security of...
Wild animals eaten by humans are known as “wild meat”, or “bushmeat” in sub-Saharan Africa. Hunting for bushmeat is both an ancient and modern practice, but as we step into the Anthropocene, bushmeat hunting has become unsustainable in many areas, threatening biodiversity and the food, financial, and cultural security of...
This project examines a number of questions using raw data on course enrollment at Duke University. Questions range from how do grades impact the order of courses students take to are there certain courses that are bottlenecks for majors. A tangible product will be a recommendation engine for course suggestion...
This project will focus on a power service restoration problem in the design of the smart grids. Due to increasingly severe weather events and cyber-physical security threats, a more resilient and reliable power system is needed to ensure the continuous operation and availability of power applications and services. Traditionally, a...
Earth observation data such as satellite imagery offers great potential for better assessing and planning for vital infrastructure systems and increasing our understanding the flow of resources in our global society. Automated analysis of such data provide their greatest benefit when they are automated and deployed at scale, being able...
Investment banking and asset management use regular market summary updates to manage risk and exposure on all equity products across markets. These Market Risk data are variable in quality based on the user who develops them. This project will aim to derive key statistical insights from Market Risk data using...
This project will apply deep reinforcement learning to solve a vehicle routing problem (VRP). VRP is a typical problem in combinatorial optimization and operations research, and it has direct applications in logistics and supply chain. A solution to VRP determines what is the optimal set of routes (in terms of...
Malicious webpages will often display brand logos to feign legitimacy; however, these brand logos are often distorted from official versions. The goal of this project is to build a classifier that recognizes and identifies logos in a screenshot of a legitimate or malicious webpage. The classifier must also be able...
In today’s world, datasets are much larger and more complex and thus require new and innovative techniques to make reliable inferences. The data also reflect existing human biases and social inequalities, which are compounded by the use of algorithms and machine learning models. This can lead to unfair outcomes and...
The goal of this project is to find a better way of presenting data from an exercise tolerance test to enable the clinician to make more precise diagnoses and better understanding results of the test. The students will create a model for this test, showing the interaction between the three...
The goal of our team was to establish an automated ETL (Extract, Transform, Load) Pipeline that extracts raw wearable data (from Fitbit and Garmin APIs), transforms the data, and loads it to the PostgreSQL database (Covidentify Analysis Database) in Microsoft Azure. This repository is a mirror of the code that...
Since the 1980s, American schools have been steadily re-segregating, despite extensive research documenting the harms of this process. This is driven by factors such wealth and opportunity which are heavily correlated with race. However, research demonstrates that the socioeconomic makeup of a school may play a more significant role in...
Every election cycle, the use of digital advertising by U.S. political campaigns becomes more commonplace. The past few years have seen extensive public debate about the use of targeted digital advertising by political campaigns. Unlike other forms of political advertising, campaign ads in the digital sphere remain largely unregulated by...
Our aim with this project was to estimate the heights of power plant stacks from publicly available satellite imagery data. Our client, WattTime, is working to build the first automated system for monitoring the emission from power plants globally, utilizing the height of flue-gas stacks to better understand the dispersal...
Our capstone project aims to develop an accurate and versatile topic modeling system. In addition, we develop a machine learning pipeline that can automatically assign labels to new articles. The project has both unsupervised and supervised components. In the first semester we accomplished the unsupervised part, where we leverage natural...
Digital libraries are an indispensable source of information in the modern era. However, given the advances in technology and the gargantuan rate of information creation, it is becoming increasingly important to enable digital libraries with intelligent solutions that aid comprehension of the constantly evolving and growing sea of scholarly publications....
Youth voted at a record rate in 2018, but of those who didn’t turnout in the last election, nearly half cited access as a reason for not voting. According to a 2019 report by The Leadership Conference, more than 1,600 polling places closed since the 2013 Shelby v. Holder decision...
Our client, Spencer Stuart, is a strategic consulting firm that offers leadership and business consulting services to organizational decision makers. Our study aims to assess S&P 500 company performance based on board member team composition. The team composition for each board includes features such as board industry experience, board female...
Saving Nature partners with local NGOs and scientists to connect fragmented ecosystems. Through land purchases and the construction of nature corridors, the organization seeks to revitalize populations of endangered species that have lost access to their historical habitats and mating grounds to industrialization. To validate the effectiveness of purchased corridors,...
With the rapid development in financial technology, algorithm trading has begun to replace human decisions. One such type of algorithm is reinforcement learning which takes trading actions by modeling historical data. In this work, we build a modular reinforcement learning system with the support of additional integrated predictive models to...
Once widely abundant throughout the North Pacific ocean, the eastern North Pacific right whale (NPRW) population (Eubalaena japonica) is now one of the rarest populations of marine mammals. Classified as critically endangered under the International Union for the Conservation of Nature, right whales have been illegally hunted for decades and...
Showing all 28 results