01 INTRODUCTION
Trusted By
Forward-Thinking Organizations
Applied across professional and collegiate sports, healthcare teams, hospitals, insurers, and more, to support high-stakes decision-making in uncertain environments.
Who We Are
Cambridge Prediction Analytics provides predictive analytics and scientific consulting to support forecasting and decision-making.
We combine Relevance-Based Prediction, mathematical modeling and computational mathematics to help organizations address complex questions with greater clarity. Our work emphasizes both the predictions themselves and an understanding of the evidence behind them.
02 METHOD
The Methodology
Relevance-Based Prediction (RBP) is a model-free forecasting framework designed
to overcome the limitations of both classical statistical models and black-box AI.​
RBP forms predictions as weighted averages of past observations, with weights determined using fundamental principles of information theory to identify the most relevant observations and predictive variables for each prediction task.
By evaluating thousands of combinations of observations and variables, RBP extracts as much information from complex datasets as AI models. Unlike AI models, however, RBP reveals exactly how each observation and variable contributes to every prediction.
03 APPLICATION
One foundation.
Two fields of application.
HEALTHCARE / CHA
Cambridge
Health Analytics
Transparent predictions for more
informed healthcare decisions.
Predictive technology for complex clinical and operational questions. Explore how relevance transparency, and prediction-specific reliability can support healthcare analysis and decision making.
Cambridge Prediction Analytics
245 Main Street, 11th Floor
Cambridge, MA 02142
Contact: prediction@cambridgeanalytics.io
Cambridge Sports Analytics and Cambridge Health Analytics are divisions of Cambridge Prediction Analytics, LLC.