Time-Series Forecasting
Time-series modeling and uncertainty-aware outputs for operational planning decisions.
Rebecca Knowlton
I'm a Data Scientist at Microsoft focused on forecasting and analytics for cloud network infrastructure capacity planning. My background includes a PhD in Statistics and industry experience in fraud detection, application health analytics, and insurance pricing across Microsoft, Apple, and USAA.
What I Work On
Risk Modeling
Statistical risk classification and threshold-based decision support for constrained resources.
Operational Analytics
Turning ambiguous infrastructure and business signals into clear, action-oriented analysis.
Industry Background
Microsoft
Data Scientist II
Forecasting and analytics for cloud network capacity planning.
Apple
Data Science Intern
Feature engineering and lift analysis for fraud decisioning.
USAA
Actuarial Analyst
Pricing analytics, catastrophe-informed adequacy reviews, regulatory filing support, and first five CAS exams passed.
Education: PhD in Statistics, University of Texas at Austin; BA in Mathematics, Trinity University.
Current Role Highlights
- Developed an interim forecasting pipeline for 180-day regional and 60-day subnet horizons with automated risk classification, including evaluation of 7 modeling approaches on 280 days of history to balance accuracy and uncertainty quality.
- Supported high-priority incident response analytics enabling rapid allocation of constrained IP capacity, followed by multi-month recovery planning support.
- Modernized inventory data quality controls with schema validation and testing, improving reliability of downstream capacity reporting and planning.
Selected Research
3 first-author publications in statistical methodology, with open-source R packages for reproducible adoption.
- Knowlton R, Tian L, Parast L (2025). A General Framework to Assess Complex Heterogeneity in the Strength of a Surrogate Marker. Statistics in Medicine, 44(5): e70001. Package: cohetsurr
- Knowlton R, Parast L (2025). Efficient Testing Using Surrogate Information. Biometrical Journal, 67(6): e70086. Package: etsi
- Knowlton R, Parast L (2026). Assessing Surrogate Heterogeneity in Real-World Data Using Meta-Learners. Journal of Causal Inference, 14(1): 20250033. Package: cohetsurr