Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Bayesia S.A.S. is a French software development company founded in 2001 by Dr. Lionel Jouffe and Dr. Paul Munteanu, specializing in artificial intelligence technology. Lionel and Paul and their team of engineers and scientists are based in Laval, France, roughly 200 miles west of Paris.
The "S.A.S." suffix in Bayesia's company name stands for "Société par actions simplifiée," which is the French equivalent of an American Limited Liability Company.
Note that Bayesia is not associated with the SAS Institute, the Cary, North Carolina-based developer of statistical software.
Bayesia's affiliates, Bayesia USA, LLC, and Bayesia Singapore Pte. Ltd. are the exclusive sales and marketing organizations for North America and Asia/Pacific, respectively. These affiliates were founded by Stefan Conrady, a German-American technology entrepreneur.
Since the expansion into Asia and the Americas a decade ago, our team has been growing the brand and successfully turned it into a synonym for Bayesian network technology.
Learn more about where you can find Bayesia around the world.
We promote knowledge discovery and reasoning with Bayesian networks to help organizations accelerate their research workflows and make better decisions.
Our flagship product is BayesiaLab, which has been under continuous development since 2001.
Today, our portfolio of research software also includes:
Bayesia Market Simulator
It's quite simple, "Bayesia" rhymes with "Asia."
Our teams in the U.S., France, India, and Singapore do offer consulting services.
We focus primarily on proof-of-concept projects for new or prospective BayesiaLab users.
We often help our customers during their first project engagement with BayesiaLab.
The revenue of Bayesia and its affiliates comes principally from software licensing fees in a B2B model.
The revenue stream is augmented by training and consulting fees.
BayesiaLab is a universal analytics platform that provides scientists with a comprehensive “lab” environment for machine learning, knowledge modeling, diagnosis, analysis, simulation, and optimization - all based on the Bayesian network formalism.
BayesiaLab supports a range of operating systems:
Windows XP, Vista, 7, 8, 10, 11 (32-bit and 64-systems)
macOS/OS X: Mountain Lion (10.8) or later (Intel and ARM)
Linux/Unix
BayesiaLab always has to be installed on your local hardware. BayesiaLab is not available as a remotely-hosted Software-as-a-Service (SaaS).
With BayesiaLab, Bayesian networks have become a powerful and practical paradigm for gaining a deep understanding of high-dimensional domains.
BayesiaLab leverages the inherently graphical structure of Bayesian networks for exploring and explaining complex problems.
BayesiaLab is employed for all kinds of research and analytics tasks.
Use cases include:
Bioinformatics
Biostatistics
Causal Reasoning/Inference
Data Mining
Demographic Analysis
Decision Analysis
Economics/Econometrics
Engineering
Epidemiology
Evidential Reasoning
Financial Analysis
Forensic Analysis
Impact Analysis
Knowledge Discovery
Knowledge Modeling
Machine Learning
Market Research/Marketing Science
Medical Research
Microarray Analysis
Operations Research
Pharmaceutical Research
Quality Analysis/Control
Reasoning under Uncertainty
Reliability Analysis
Risk Analysis
Root Cause Analysis
Sports Analytics
Strategic Decision Support
We see BayesiaLab as a complementary tool with regard to more traditional, parametric statistical methods.
BayesiaLab really shows its strength when it comes to complex, high-dimensional problem domains and reasoning under uncertainty.
After completing the Introductory BayesiaLab Course, it typically takes six months to one year of practical use to become proficient in using the software platform.
Embracing BayesiaLab as a research framework requires a significant commitment in terms of time and resources.
BayesiaLab's data capacity is only limited by the RAM of the computer on which it is installed.
All of BayesiaLab's algorithms are performed in memory.
Furthermore, BayesiaLab offers database connectivity via JDBC.
BayesiaLab Professional is not available as SaaS.
However, BayesiaLab can be installed on your own virtual machine, e.g., on an Amazon EC2 server.
For setups with multi-user access, the respective license fees apply.
BayesiaLab and the Bayesia APIs always run locally on the hardware on which you installed the software, e.g., your desktop or laptop computer.
There is no provision for running BayesiaLab remotely on Bayesia's servers.
Your local BayesiaLab installation never transmits any of your data or models to the Global Bayesia License Server.
The only information sent to the Global Bayesia License Server is the user login and hashed password, the user account, and the version of BayesiaLab software currently in use.
For each session, only the username, account, start and end date/time of the session, plus the associated IP address, are stored on the Global Bayesia License Server.
Any temporary connection ("ping") to the Global BayesiaLab License Server only serves the purpose of validating your licenses.
All data and models are stored within the infrastructure you control unless you explicitly choose to upload them elsewhere, for instance, to the BayesiaLab WebSimulator Server, with the express intent of sharing your model.
Bayesia staff has no access to or visibility of your models or data through the BayesiaLab software.
You can access many of BayesiaLab’s functions outside the graphical user interface by using the Bayesia Engine APIs.
They allow developers to integrate BayesiaLab’s technology into their own applications.
The APIs are available as:
Java API
REST API
Models with a Target Node, i.e., those generated by supervised learning algorithms, can be exported as code for
SAS
PHP
JavaScript
VBA
Python
R.
This Code Export Module is available as an add-on module to BayesiaLab by subscription.
License Rental (3 months, 6 months, 12 months)
License Purchase (perpetual) plus annual maintenance
Elastic Licensing
Please see our pricing pages for details on license configurations and pricing.
A fully-featured Academic Edition is available at a substantial discount.
Additionally, we offer a BayesiaLab Education Package, which provides an economical way of using BayesiaLab for classroom instruction.
A Bayesian network, also known as a Bayesian belief network, is a probabilistic model representing a set of random variables and their conditional dependencies using a directed acyclic graph (DAG) and a set of conditional probability distributions.
In a Bayesian network, each node in the DAG represents a random variable, and each directed arc represents a conditional dependency between the variables. The nodes in the network are associated with probability tables that specify the probability distribution of each variable given the values of its parent variables.
Bayesian networks are used to model complex systems and make predictions based on incomplete or uncertain information. They can be used for various tasks, such as classification, prediction, diagnosis, decision-making, and causal reasoning.
One of the critical advantages of Bayesian networks is that they allow for the efficient representation and computation of complex probability distributions. They are particularly useful when relationships between variables are complex and difficult to model using traditional statistical methods.
Bayesian networks provide an elegant and sound approach to represent uncertainty and carry out rigorous probabilistic inference by propagating the pieces of evidence gathered on a subset of variables on the remaining variables.
Bayesian networks are not only effective for representing experts’ beliefs, uncertain knowledge, and vague linguistic representations of knowledge via an intuitive graphical representation but are also powerful knowledge discovery tools when associated with machine learning and data mining techniques.
From a technical point of view, a Bayesian network consists of two parts:
Qualitative: a Directed Acyclic Graph (DAG), i.e., a special kind of directed graph that does not include cycles.
Directed Acyclic Graphs are composed of nodes that represent the variables of the domain (e.g., the temperature of a device, a feature of an object, the occurrence of an event, the age of a patient), and the links represent statistical (informational) or causal dependencies among the variables.
The DAG is the formal definition of the factorization of the Joint Probability Distribution over the set of all variables in the domain.
Quantitative: conditional probability distributions to quantify the dependencies of each node given its parents in the DAG.
Let’s take two variables: Age and Gray Hair. The corresponding DAG has two nodes, one for Age and another for Gray Hair.
As there is a probabilistic (and causal) relationship between Age and Gray Hair, there is a link between the two nodes:
The probability distributions in a Bayesian network are typically represented with tables. The marginal distribution of the node Age is illustrated in the following table:
This table tells us, for instance, that 16.4% of the population under study is less than 30 years old, while 9.4% is more than 70 years old.
The following Conditional Probability Table quantifies the relationship between Age and Gray Hair:
This suggests that among those under 30 years of age, 66% do not have any gray hair. Conversely, among people older than 70, 30.8% are completely gray.
The DAG and the probability distributions associated with each node allow a compact representation of the Joint Probability Distribution over all the variables.
Inference algorithms are available so that Bayesian networks can be used as probabilistic expert systems or inference engines for computing the posterior probability distributions of unobserved nodes given evidence observed on any number of observed nodes.
Also, observational inference in Bayesian networks is omnidirectional: it is possible to perform inference from parent nodes to child (simulation), child nodes to parent nodes (diagnosis), and any combination of the two kinds of inference.
However, it is essential to point out that causal inference can only be used in the context of simulation with a causal Bayesian network.
The following Monitors show the marginal probability distributions of Age and Gray Hair:
Simulation Example
In this example, we simulate the posterior probability distribution of Gray Hair, given that we observe Age>70.
The top Monitor shows the observed evidence (Age>70); the bottom Monitor displays the posterior probability distribution of Gray Hair.
Diagnosis Example
Now, we reason in the reverse direction. We observe Gray Hair=Moderate and infer (or diagnose) the posterior probability distribution of Age.
There are two ways to create a Bayesian network:
Knowledge Modeling: You can use any available expert knowledge to manually design a Bayesian network and define the corresponding probability distributions.
Machine Learning: You can machine-learn a Bayesian network from data and estimate the corresponding probability distributions.
Within the same theoretical framework, BayesiaLab offers a broad set of data mining algorithms:
Unsupervised Learning: BayesiaLab induces a Bayesian network to compactly represent the joint probability distribution sampled by the data set; all the variables have the same importance in this context.
Supervised Learning: BayesiaLab can learn a Bayesian network entirely focused on the characterization (or prediction) of a target variable.
Data Clustering: BayesiaLab creates a Bayesian network with a hidden variable to represent uniform groups of individuals/observations.
Variable Clustering: BayesiaLab identifies strongly connected variables that can be clustered into factors.
Probabilistic Structural Equation Models: BayesiaLab builds a hierarchical Bayesian network using hidden variables (or factors) that were identified during Variable Clustering.
This graph defines a factorization of the joint probability distribution :
BRICKS is a Probabilistic Relational Modeling framework that supports creating and utilizing models of large-scale and complex systems.
The BRICKS methodology consists of three distinct phases:
The knowledge modeling phase produces a generic knowledge base of BRICKS classes.
The system design produces the model itself by instantiating the previously created BRICKS classes.
The querying phase lets you interact with the model by providing observations and requesting probability distributions.
To learn more, please see the dedicated BRICKS website: bricks.bayesia.com
BEST provides a universal modeling framework based on a powerful AI for troubleshooting complex systems (composed of interrelated and/or heterogeneous components: mechanical, electric/electronic, software) that exhibit symptoms depending on their configurations and their failure modes:
Deteriorated performance
Failure codes,
Error messages
Warning lights
Abnormal noises or leaks,
Test results
Unsuccessful repair attempts
Etc.
To learn more, please see the dedicated BEST website: best.bayesia.com
The Thinker of Hamangia
The Thinker of Hamangia is a famous Neolithic sculpture discovered in the Hamangia culture of Romania. It is a small statue made of baked clay dating back to around 5,000 BC, making it one of Europe's oldest examples of sculpture. The statue depicts a seated figure with its chin resting on one hand, suggesting a contemplative or thoughtful posture, hence its name. It is considered a significant artifact in understanding the art and culture of prehistoric Europe.
The statue depicts a figure in a contemplative pose, which symbolizes deep thought and intellectual exploration—concepts closely associated with Bayesia's innovations in Artificial Intelligence. Additionally, the statue pays homage to the home country of Dr. Paul Munteanu, one of Bayesia's co-founders.
The figure in the Bayesia logo represents the Thinker of Hamangia.
Parc Ceres, Batiment N 21, rue Ferdinand Buisson 53810 Change, France
Even though Bayesia's R&D team is fully occupied with BayesiaLab's development roadmap, everyone in the organization also spends time on consulting projects with clients. Dr. Jouffe says, "We must be intimately familiar with our customers' evolving research challenges so the tools we develop fit into their workflows. We are driven by the idea that software should naturally anticipate the user's next steps and yet offer a maximum degree of flexibility to the researcher. After all, research is a voyage of discovery and not a script."
4235 Hillsboro Pike Suite 300-688 Nashville, TN 37215
In his capacity as head of analytics and forecasting at Nissan North America, Stefan Conrady started collaborating with Bayesia in 2009 on market research applications of BayesiaLab in the auto industry. Convinced of BayesiaLab's tremendous potential beyond its home market in Europe, Stefan founded Bayesia USA in 2010 to serve as the French company's North American sales and marketing partner.
In their first year of operation, Bayesia USA already managed to sign up a number of blue-chip clients, such as NASA's Jet Propulsion Laboratory, InterContinental Hotels Group, and Dell Computers, and add them to the expanding list of BayesiaLab users around the world. Bayesia USA has also become the center for educational activities around BayesiaLab, hosting webinars, workshops, and training seminars around North America. The 382-page book Bayesian Networks & BayesiaLab — A Practical Introduction for Researchers, co-authored by Stefan Conrady and Lionel Jouffe, is the latest product of Bayesia's international teamwork.
Stefan's work experience with some of the world's leading automobile brands, including Mercedes-Benz, BMW, and Rolls-Royce Motor Cars, instilled in him a sense of commitment to total customer satisfaction. "If the customer runs into a roadblock, for whatever reason, our team will try the utmost to solve the problem." He adds, "Research is expensive, and BayesiaLab is a big technology investment, so our clients can't afford any downtime. For them, it's critically important to be able to rely on our industrial-strength support."
1 Fusionopolis Place #03-20 Galaxis Singapore 138522
Located at the hyper-modern Fusionopolis complex, Bayesia Singapore is at the epicenter of Singapore's research and development activities. Situated in Buona Vista, Fusionopolis is close to the ESSEC and INSEAD Asia Campuses, the National University of Singapore, the Singapore Polytechnic, the Institute of Technical Education, the National University Hospital, the Singapore Science Park, the Biopolis, and the Ministry of Education.
Beyond our corporate presence on three continents, our reach through courses and seminars goes much further. Here is a map of the cities where we host courses and seminars on a regular basis.