Evan B
- Research Program Mentor
None candidate at
Expertise
Data Science, Machine Learning, Bioinformatics, and Public Health
Bio
To quote one of my research mentors, we have to “make our data work for us”. I have been working in academic research and data science for over three years. Having graduated with a B.S. at Northeastern University majoring in Biology and minoring in data science I have learned how data can solve really important problems ranging from using machine learning to fuel novel drug discovery to using data to identify how a certain policy has a negative impact on a community. Currently, I am working at Boston Children’s Hospital and Harvard Medical School as a research assistant. In my current position, I lead exciting projects using datasets containing tens of millions of patients to better understand health disparities. Previously, I have completed a publication using image-based neural network models to detect presence of pediatric appendicitis using ultrasound images from a patient. If you are interested in using data, to help solve and address problems; I will help guide you to craft your ideas in interesting directions and to help you tackle hard problems into easy, manageable steps. Working with “big data” and asking interesting questions may intimidate you, but I will show you how to use large datasets to help fuel your research interests. In my free time, I love reading books. Specifically, my favorite author is Robert Caro. I think about Robert Caro's books and draw references to them daily.Project ideas
Social Justice Engagement Project
Social justice data science projects can allow you to think “globally, but act locally”. For example, if there is a new crime law in your community determining, using a dataset released from your community, can help portray whether this new law has positively affected your community or not. This would require using data visualization graphs to report findings in an interesting and interpretable way. You will also use fundamental statistical tests to validate your results. If the findings are interesting, the findings can be written about in a blog post and/or be reported to an elected official in your community.
Detecting Public Opinion
Using text data from Twitter one can model how certain information is represented and spread to other Twitter accounts. An example of this is modeling how public opinion changes over a period of time. Specifically, to model the sentiment over six months regarding whether the trend of people who were for and against children running to in-person schooling during the Covid-19 pandemic.
The Good and the Bias of Artificial Intelligence
Artificial intelligence is increasingly more and more in our lives. For these projects you can create a machine learning model as a practical utility. Such a machine learning can be developed using text, image, or other variables. These projects can range from using radiology images to predict patient diagnosis to developing solutions to help prevent existing machine learning models to perpetuate social biases into a model’s prediction. I will help guide you through what goes into building a novel dataset, if needed, then to train and test the model to ensure the model perform like you intended.