Introduction

Bayesian Belief Networks (BBNs) have been used to model causal and probabilistic reasoning across several industries and problem domains. It is the de facto tool of choice for answering questions around causality. While a BBN is natural and beautiful to use to help decision and policy makers in their roles, learning one is very difficult. In general, given any number of variables under investigation, there is an exponential number of BBNs that may fit and/or explain the data.

The Art of the Possible

With pyspark-bbn, learning a useful BBN to solve a problem is within reach. The key differentiator of pyspark-bbn is its cloud-native, cloud-first approach to learning BBNs over massive and high-dimensional data. We leverage the popular Apache Spark framework for demanding computational needs. To our knowledge, pyspark-bbn, is the first of its kind leveraging cloud computing frameworks to learn BBNs.

Simple, unified and effective

The objective of pyspark-bbn’s technical design is to provide a simple, unified and effective framework for learning BBNs. Under the rapid changes of the world economy and global events, decision and policy makers need a way to understand causally and probabilistically the impacts to their business. When they and/or their team try to understand how to best prevent or intervene to mitigate problems or provide alternative solutions, the last thing desired is a formidable scientific and technical hurdle. Thus, pyspark-bbn will always have the underlying philosophical design of ease-of-use while maintaining rigorous and disciplined computer science implementations.

Products and Service

Let us help you achieve high-value, high-impact changes to your organization or problems. We are the original authors of py-bbn, an exact inference API for BBNs, and now we bring you pyspark-bbn. We go beyond products and also provide services to educate you, your team and organization to go beyond prediction and explanation, and towards causation. We also specialize in services to help your data science practice adopt and use causal and probabilistic tools such as BBNs to solve your most pressing business problems. Give us a chance to bring order-of-magnitude changes to your business and problem domain using our expertise.