Belief Propagation Algorithm for Complex Networks

Quite new to me,   see my link to 'belief propagation' ... which points to MIT work on Kalman Filters.  taking a closer look.    We worked with SFI.  

Can the 'Belief Propagation' Algorithm Accurately Describe Complex Networked Systems?  By Santa Fe Institute,  April 29, 2021

Researchers at the University of Michigan and the Santa Fe Institute (SFI) demonstrated a novel belief propagation algorithm to solve probabilistic models on networks containing short loops.

These algorithms can be used to model the spread of a disease, for instance, by looking at people in close contact with each other, not their entire network.

However, SFI's George Cantwell said, "Suppose Alice was in close contact with Bob, who was in contact with Charlotte. To know what happens to Alice, we need to know about Bob, and then Charlotte. But suppose it turns out that Charlotte was already in contact with Alice, now we've backed ourselves into a sort of infinite regress. To predict what happens to Alice, we need to first predict what happens to Bob, then Charlotte, then Alice again."

The researchers showed their method could make accurate theoretical predictions for realistic networks.

From Santa Fe Institute

  .. A messaging-passing algorithm known as belief propagation can be used to analyze large systems by breaking them down into smaller pieces and ensuring all the smaller solutions are consistent with each other. ... 

MIT Model Forecasts Business Financials

If real, here is something of considerable value ... Note 'Belief Propagation' .... which we tested for things like predicting future product sales in varying context.  Looking more closely.

Model Beats Wall Street Analysts in Forecasting Business Financials
MIT News
By Rob Matheson

Massachusetts Institute of Technology (MIT) researchers have developed an automated model that significantly outperforms humans in predicting business sales using only anonymized weekly credit card transactions and three-month earnings reports. The researchers used the model to predict quarterly earnings of more than 30 companies, and found that it outperformed the combined estimates of expert Wall Street analysts on 57% of predictions. The human analysts had access to any available private or public data and other machine learning models, while the MIT model used a very small dataset of the two data types. The researchers used a variation of the standard inference algorithm, called Kalman filtering or Belief Propagation. This technique uses data measurements observed over time, containing noise inaccuracies, to generate a probability distribution for unknown variables over a designated timeframe .... "