Bryant University values hands on learning, but also makes strides in research. This past summer there have been teams working in microbiology research under the Summer Undergraduate Research Fellowship (SURF) program. Professors Jiang and Primus and Instructor Ratcliffe have made strides in creativity and industry disruption, which they have presented to the International Society of Professional Innovation Management (ISPIM) conference. Associate Professor of Marketing Sukki Yoon, Ph.D., has also been published in the Journal of Advertising and has received a grant from the National Research Foundation of Korea to extend his research.

Michael Gough | Class of 2018 | Finance, Economics, Applied Statistics

Research Project: “Can the Put to Call Ratio Improve Existing Asset Pricing Models?”

Michael Gough has recently been working on research for his Honor’s Capstone Project. The original research question was “Can the Put to Call Ratio Predict Equity Returns?” but the initial results were somewhat disappointing and his advisor, Professor Inci, pivoted to form the question “Can the Put to Call Ratio Improve Existing Asset Pricing Models?” After adding the put to call ratio as a sentiment indicator, things became more interesting as they have seen early signs of model improvements. Before making any serious claims they would like to expand their test set from ten stocks to all Dow 30 stocks.

For his analysis, Michael used Python and the Pandas, StatsModels, and Matplotlib packages. Pandas for data manipulation and cleaning, StatsModels for tests, and Matplotlib to create visualizations. Michael wrote all of his code in a Jupyter notebook because you can easily make markdown comments to improve readability.

Michael believes and encourages all students to conduct their own research. Most undergraduates at prestigious universities write a senior thesis. Given the chance he would encourage all Honors Students, in particular, to stick with the program for the support network and resources. If you’re not in the Honors Program, no worries, you can find a professor with similar interests and propose a directed study. Be a life-long learner. Next semester Michael will be working to investigate fixed income markets.

Anthony Pasquarelli | Class of 2018 | Information Technology, Applied Analytics

Research Project: “Feature Detection in Medical Images using Deep Learning”

Anthony has been collaborating with individuals from the Rhode Island Hospital who work in the Computer Vision and Image Analysis Lab. This is where they search for ways to leverage the huge amounts of data and images that get collected from X-Rays, CT scans, and MRIs. They look for new ways to use image analysis to aid diagnosis, prevent errors, and improve patient care.

Anthony has been using Keras, which is a high-level machine learning platform that uses TensorFlow as a backend. Keras allows him to build convolutional neural networks from scratch or import existing network architectures, then train them to look at images and make predictions. So far, he has been utilizing the InceptionV3 network, which was developed by Google, to predict the age of children just by looking at X-Rays of their hands. After training on over 12000 images multiple times, the network is able to predict age with an average error of 8 months.

Anthony plans on refining this process to reduce the error margin and also look at different types of problems, most likely dealing with 3D data from CT scans or MRIs. One potential area of interest is the detection of cancer nodules in lungs. Anthony advises all Data Science majors at Bryant to learn to code and be as technical as possible. He believes it will increase your value to employers significantly. Anthony would also like to thank Professor Blais for agreeing to be his capstone advisor.

Jake Schurch | Class of 2018 | Finance, Computer Information Studies, Applied Analytics Research Project: “Machine Learning in Robotics”

Jake Schurch recently interned with Ernst & Young with their advisory team in FSO and has accepted an offer for a fulltime position. He has mentored many students around Bryant about the success that comes from learning to code in Python to in order to gauge their interest in emerging fields such as FinTech. Many industries are experimenting and implementing machine learning solutions in order to drive desired factors. A good example of is AI-driven financial investing and driverless cars. His directed study revolves around using the python programming language. Specifically, he used Professor Blais’s custom game package to implement AI in a board game; and is now using Google’s TensorFlow object recognition package to detect common objects in images.

What he has been able to deduce in his directed study is that most Machine Learning / AI solutions revolve around pattern-matching. Although the patterns these models use are much more complex than any patterns a human can match, it is still pattern-matching nonetheless; which he thinks is extremely interesting. Jake hopes to utilize machine learning in future coding projects and would like to thanks Professor Blais for making his directed study extremely rewarding.

All three students in the article are open to any questions. They recommend that anyone who might be interested in doing similar applications in research should definitely reach out. To those interested in research, start inquiring and asking professors early on in your undergraduate career, and maybe one day your research can and will change the world.