Bridget Bertoni, Phd Email & Phone Number

- Created the study population for Medicare members with falls and developed models to predict initial and subsequent fall risk as well as clustering models to identify clinical profiles to inform intervention design, using Pyspark Azure databricks- Applied model interpretability methods (SHAP, LIME, EBM) and causal inference models (econML) to evaluate interventions and for subgroup analysis- Improved and automated Humana vs industry utilization and cost benchmarking- Developed… Show more - Created the study population for Medicare members with falls and developed models to predict initial and subsequent fall risk as well as clustering models to identify clinical profiles to inform intervention design, using Pyspark Azure databricks- Applied model interpretability methods (SHAP, LIME, EBM) and causal inference models (econML) to evaluate interventions and for subgroup analysis- Improved and automated Humana vs industry utilization and cost benchmarking- Developed data-driven models to predict disability and ESRD status for Medicare members- Designed flexible Power BI dashboard connected to live data in Azure Synapse Analytics SQL pools to explore IP admissions and LOS using DAX and Python Show less

- Led discussions with customers and subject matter experts to determine data science solutions with the highest business impact- Led the data science effort to create and operationalize an ML-based search algorithm over 4 million documents stored in a SQL DB using TF-IDF, Word2Vec, and serverless compute- With a team, designed and operationalized a search app that used machine learning built on top of Azure Cognitive Search to extract key text and classify hundreds of thousands of… Show more - Led discussions with customers and subject matter experts to determine data science solutions with the highest business impact- Led the data science effort to create and operationalize an ML-based search algorithm over 4 million documents stored in a SQL DB using TF-IDF, Word2Vec, and serverless compute- With a team, designed and operationalized a search app that used machine learning built on top of Azure Cognitive Search to extract key text and classify hundreds of thousands of scanned manufacturing documents- Built an XGBoosted Gradient Tree model to predict course enrollment for college classes based on historical enrollment and new course recommendations Show less

Burlingame, Ca

- Developed payment models to forecast Medicare spending for sepsis - Proposed needed changes to the two-equation regression model for Medicare's $12 billion ESRD payment system (ESRD PPS) to improve accuracy and prevent gaming of the payment system- Redesigned and automated payment forecasting code to set payment rates and analyze impact- Constructed patient timelines and executed discrete-time survival analysis to predict disenrollment- Combined survey data sets and Medicare… Show more - Developed payment models to forecast Medicare spending for sepsis - Proposed needed changes to the two-equation regression model for Medicare's $12 billion ESRD payment system (ESRD PPS) to improve accuracy and prevent gaming of the payment system- Redesigned and automated payment forecasting code to set payment rates and analyze impact- Constructed patient timelines and executed discrete-time survival analysis to predict disenrollment- Combined survey data sets and Medicare data to understand the demographics of populations eligible for but not enrolled in various Medicare programs- Designed and managed a project for a summer intern Show less

- Utilized Monte-Carlo methods, Python, and bootstrapping to calculate correlation functions in Milky Way stellar data and simulated data to search for dark matter substructure- Initiated original research to study gravity wave signals to search for dark matter

- Applied quantum field theory in novel scenarios to constrain parameters in dark matter models- Solved non-linear equations in Mathematica and Python to find and visualize mathematical results

- Graded homework and exams and substitute-lectured for undergraduate and graduate physics courses in statistical mechanics, electricity & magnetism, quantum mechanics, and particle physics- Taught introductory physics labs and problem solving/discussion sections

- Wrote Python and shell scripts to analyze data from the Fermi Large Area Telescope (describing thousands of astrophysical objects) and found statistically significant hints of dark matter- Evaluated statistical significance of model fits using Poisson statistics and a log-likelihood analysis

Master'S Degree (2012), Phd (2015), Theoretical Particle Physics

Activities and Societies: Mentor in the Physics Mentoring Program, Co-organizer of the Particle Theory Journal Club

Bachelor Of Science (B.S.), Physics And Mathematics