Venkata Sukumar Gurugubelli

January 2018 - Current

University of Massachusetts - Dartmouth

Doctoral Researcher

• A thought leader in the field of Data Science and Machine Learning
• Designed and developed a novel neuro-fuzzy machine learning model (R-mGNNF) for outcome prediction in extensive multi-dimensional longitudinal data using PyTorch, Tensorflow, and Jupiter Notebook, which improves the accuracy and computational performance over traditional predictive models.
• Developing visualization-based validation methodology for R-mGNNF to improve the interpretability and traceability of neuro-fuzzy models.
• Implemented various predictive modeling techniques such as logistic regression, decision trees, random forests and naïve bayes for outcome prediction of clinical data.
• Innovated novel Support Vector Machine (SVM) based regularization approach for improving precision and recall of neuro-fuzzy predictive models.
• Harmonization of complex high dimensional longitudinal dietary data from six major study centers with more than 200k rows and 200+ diet history and demographic variables using SAS, SQL, SPSS, Python, and shell programming as part of one of the most significant harmonization initiatives funded by the National Institute of Health (NIH).
• Preprocessed big data and performed statistical analysis using dplyr, ggplot2 in R, SAS, SPSS, Python, and shell programming for drawing insights from complex raw clinical data.
• Conducted Exploratory Data Analysis on intervention and observation study data to draw insights from data using Python, R and SAS.
• Led multiple research projects and mentored junior research members in research methodologies.
• Utilized tools such as Tableau and D3 as needed for presenting data insights and visualizations.
• Collaborated with researchers from 4 local and two national study centers to reach grant goals in a timely fashion.
• Published research work at data science related conferences and Journals as part of PhD program.

June - Sept 2017

UMass Chan Medical School

Research Assistant

• Collaborated with three computational scientists to design, develop, test, and validate machine learning models using real longitudinal data.
• Developed a novel fuzzy-based clustering model (eFCM) to predict the outcome of smoking intervention data to improve accuracy and computational performance over traditional clustering models.
• Planned and organized a research plan to develop a novel fuzzy clustering algorithm and spear-headed research to fulfill PI's goals.
• Published paper and presented at an international conference in collaboration with two colleagues.

May 2013 - June 2015

Contenterra Software Pvt. Ltd.

User Experience Researcher

• Improved User experience by providing additional means to contact developers by collecting metadata of the application's page from where the user is reporting the bug. The impact was the reduced time for a bug report to reach the developers of features on a page.
• Promoted as User Experience Designer to focus on fine-tuning the interfaces by reiterating interactions across the application where necessary, analyzing user feedback, and conducting quantitative analytics using python and the data obtained from user surveys.
• Coordinated with a team of 4 developers, three testers, the manager, and the principal of the company when making decisions and implementing suggested improvements to user interfaces.