M.S. in Data Science


UNCW Data Science Practicum

The UNCW Data Science Program strives to train students in technical proficiency as well as business acumen. While course work  and research projects are sufficient to satisfy the first goal, a team-based professional experience, with organizations external to the university, is preferred to meet the second goal.  In order to achieve this, the program seeks partnerships with industry, non-profits, and other groups to provide such an experience in the form of a Practicum.

The Practicum is a project sponsored by an external organization that involves teams of 4-5 students working on a data analysis problem posed by the sponsoring organization. The project spans eight months and is co-managed by the sponsoring organization and the UNCW Data Science faculty. The financial commitment of sponsoring organizations is limited to any potential travel costs for students to visit their organization for a kick-off meeting.

More detailed information is available in the Practicum Guide and the Sample Nondisclosure Agreement(NDA) .

To submit an idea for a Practicum please fill out this Practicum Proposal Form

2023  Data Science Practicum Timeline

  • Call for Proposals - February 8th
  • Initial Proposal Review/Proposal Submission Deadline - March 24th
  • Practicum Sponsors selected - April 7th
  • NDA agreements signed and Data Security finalized - April 24th
  • Practicum kick-off meetings - May 3rd - 9th
  • Summer - conference and video calls at convenience of sponsor
  • Bi-weekly meeting with team and sponsor beginning August 23rd
  • Midpoint review meeting by October 1st
  • Final Presentations by Dec 11th


2022 Practicum Partners


In partnership with GE Hitachi, Investigated the application of data science algorithms and methods to critical power correlation.


Partnered with Live Oak Bank  to gain a deeper understanding of how economic conditions impact production and churn across various products and improve organizational decision making.


Used NLP to leverage posts on nCino’s social media platform. Built insights and developed strategies to help optimize company’s platform.


800px-SAS_logo_horiz.svg.png  Lightcast.png  

Analyzed sales data to give factors that may help predict if an institution is going to purchase a product and when the dela is likely to close. 


2021 Practicum Partners

 CFC Logo

Cape Fear Collective and Costal Connect Health information exchange. Hospital
utilization and comorbidity analysis during the Covid-19 pandemic.

Defense Storm

Development of timeseries methods and machine learning algorithms for anomaly detection of global cyber security threats for outsourcing financial institutions.


Partnered with IBM, to use the PAIRS dataset for various projects. This data ranges from COVID19 measures to weather, all available spatio-temporally. The work  required pulling data from IBM’s database, manipulating and combining the data set(s), creating summary assessments and visualization of the data, and performing statistical analysis. alyzed


The team used Spacy, BERT, Regex, and data profilers to train a named entity recognition model to find identifiable information on financial documents




2020 Practicum Partners

GEHitachi.PNG GE- Hitachi

• Used data from the US Energy Information Administration to build models for consumption and
cost of energy used to generate electricity in the US.
• Created time series models as well as other approaches to make projections about cost
effective strategies for future electrical consumption planning

Live_Oak_Bank.jpg Live Oak Bank

• Analyze customer call data for Live Oak Bank to optimize their customer experience and incorporate metrics into dashboards and other monitoring tools
• Use sentiment analysis on reviews for Live Oak Bank to find out what customers specifically like and dislike about the bank to improve their experience
• Use time series analysis to forecast the number of calls in the future to anticipate the need, if any, to hire more employees to meet a growing demand

GEHitachi.PNG GE- Hitachi

• Improving the Fidelity of the adaption process in boiling water nuclear reactors.
• Data analytics and machine learning were performed on the traversing in-core probe instrumentation
alignment data to have an assessment for a goodness metric.


• Collect and clean data involving recalled pet products from U.S FDA Recall and Withdrawals
• Investigate and analyze studies on key contamination recalls

GEHitachi.PNG GE- Hitachi

• Scope of project includes ingestion of raw boiling water reactor core-monitoring data, application of Unsupervised Learning Machine Learning methods to detect anomalous power and/or ow conditions, and mockup for result dashboard.




2019 Practicum Partners

CastleBranch.PNG Castle Branch

Students cleaned and merged data to create a database used for reporting and visualizations.
They used the data to find trends in operations, sales and employee efficiency.

GEHitachi.PNG GE- Hitachi

Students modeled and analyzed supply chain behaviors and/or processes using quantitative data modeling to align inventory-purchasing-manufacturing schedules to optimize cash. They used predictive analytics to forecast bill of materials demand and forecasts to minimize costs associated with orders.

GEHitachi.PNGGE- Nuclear Fuel

Students  worked on improving the fidelity of the adaption process in boiling water nuclear reactors. Data analytics and machine learning were performed on the traversing in-core probe instrumentation alignment data to have an assessment for a goodness metric. 


Students worked on  developing a pet food recall database using data from the FDA, provided insight on what trends may exist in the data and began developing models to help predict future recalls.

combomark_primary_sm.png  Prometrics

Students identified Oncology patient populations from  real world Medical and Prescription claims data sets using industry standard diagnosis, procedure, and drug codes. They developed methods to automatically identify distinguishing characteristics of a patient population through feature selection, etc.

Tillium.PNGTrillium Health

Students developed predictive analytic methodology using medical data. They predicted elevated risks of hospitalization and clinical intervention for patients with a serious and persistent mental illness diagnosis and were not taking psychotropic medications.


2018 Practicum Partners

 Live_Oak_Bank.jpg Live Oak Bank

Students performed time series forecasting of bank liabilities including savings, cd and money market accounts.  

 nCino_Logo-Full_color-Light_bg.png nCino

Students assisted nCino in automating banking spreads technology to reduce manual input and processing time through machine learning and text classification methods.

combomark_primary_sm.png Prometrics-

Students manipulated and analyzed medical claims data to find the optimal time-to-delivery post diagnoses of disease specific pharmaceuticals.

IRP_UNCW.PNG UNCW Office of Institutional Research and Planning

Students usee SAS 9.4 to create an interactive dashboard analyzing and visualizing teaching space utilization at UNCW. Not under NDA so  product may be viewed here Campus Map Room Use.