www.bayesia.com does not fully support your browser (Internet Explorer 6).
We suggest upgrading to IE 7 or downloading Firefox for a more enjoyable web experience.

Newsletter - November 2011

November, 17 2011

  BayesiaLab Training Sessions

The next training seminars "Bayesian Networks for Expert Knowledge Modeling and Data Mining" will take place in:

• Laval, France, December 06-08 (in English), registration deadline December 01
• Paris, France, January 17-19 (in French)
• Orlando, FL, USA, February 13-15 (right before the ASA Conference on Statistical Practice)
• Paris, France, March 20-22 (in French)
• Orange County, CA, USA, April 18-20 (right after the INFORMS Conference on Business Analytics & O.R.)
• Paris, France, June 19-21 (in French)
• Toronto, Canada, July 10-12
• Paris, France, September 25-27 (in French)
• San Jose, CA, USA, October (to be held in conjunction with the 2012 Data Mining Camp)

We cover all the fundamentals of Bayesian Networks and Data Mining, so no prior knowledge is required other than a basic familiarity with mathematical and statistical concepts. In conjunction with the seminar, participants have access to a 60-day unrestricted license of BayesiaLab Professional Edition, so they can experience the full array of features through the hands-on exercises.

Click here for details about the training and registration information »

White Papers

Here is the list of our most recent White Papers available in the BayesiaLab resources section:

• Introduction to Bayesian networks
  Bayesian networks are somewhat of a disruptive technology, as they challenge a number of common practices in the world of business and science. So, beyond the world of academia, promoting Bayesian networks as a new tool for practical knowledge management and reasoning still requires significant persuasion efforts. With this short paper, we attempt to provide a concise justification, both from a practitioner's and a technical perspective, why Bayesian networks are so important.
  
  
• Paradoxes and Fallacies: Resolving some well-known puzzles with Bayesian Networks
  There are a number of paradoxes and fallacies that keep recurring as popular and mind-bending puzzles in the media. Although there is (now) complete agreement among scientists on how to resolve them, the correct answers are often perplexing to the casual observer and still cause bewilderment. As this paper will show, the formally correct solutions of these probabilistic paradoxes are counterintuitive.
  
  
• Driver analysis and product optimization with BayesiaLab
  This tutorial provides hands-on examples of how to use Bayesian networks in the field of marketing science. In this study we want to examine how product attributes perceived by consumers relate to purchase intention for specific products. Put simply, we want to understand the key drivers for purchase intent. Given the large number of attributes in our study, we also want to identify common concepts among these attributes in order to make interpretation easier and communication with managerial decision makers more effective. Secondly, we want to utilize the generated understanding of consumer dynamics, so product developers can optimize the characteristics of the products under study in order to increase purchase intent among consumers, which is our ultimate business objective.
  
  
• Probabilistic Latent Factor Induction and Statistical Factor Analysis
  It is not surprising that the new Bayesian network paradigm prompts comparisons to more conventional methods. In the field of market research, for instance, long-established methods, such as factor analysis remain in daily use today. Given that there exists a direct counterpart to factor analysis in the Bayesian network framework, we want to highlight similarities as well as fundamental differences. The objective of this paper is to present both methods side-by-side and thus help researchers to correctly compare and interpret the respective results. More specifically, we want to establish the semantic equivalents between the traditional statistical factor analysis approach and BayesiaLab's method based on Bayesian networks, which we refer to as Probabilistic Latent Factor Induction.
  
  
•  Causal Inference and Direct Effects
  The format of this document is essentially "two papers in one," with the first chapter focusing on mostly theoretical considerations (although illustrated with an example), while the second chapter provides a practical, real-world example presented in the form of a tutorial. 
  
  Methods of Causal Inference: We will first introduce the reader to the idea of formal causal inference using the well-known example of Simpson's Paradox. Secondly, we will provide a brief summary of the Neyman-Rubin model, which represents a traditional statistical approach in this context. Once this method is established as a reference point, we will introduce two methods within the Bayesian network paradigm, Pearl's Do-Operator, which is based on "Graph Surgery", and a method based on "Likelihood Matching" algorithm (LM). LM allows fixing probability distributions and can be considered as a probabilistic extension of statistical matching.
  
  Practical Applications of Direct Effects and Causal Inference: While our treatment of Neyman-Rubin is limited to the first chapter, the two Bayesian network-based methods will be further illustrated as practical applications in the second chapter. Special weight will be given to Likelihood Matching (LM), as it has not yet been documented in literature. We will explain the practical benefits of LM with a real-world business application and discuss observational and causal inference in the context of a marketing mix model. Using the marketing mix model as the principal example, we will go into greater detail regarding the analysis workflow, so the reader can use this example as a step-by-step guide to implementing such a model with BayesiaLab.
  
  
• Unsupervised and Supervised Learning with BayesiaLab
  Perhaps more than any other kind of time series data, financial markets have been scrutinized by countless mathematicians, economists, investors and speculators over hundreds of years. Even in modern times, despite all scientific advances, the effort of predicting future movements of the stock market sometimes still bears resemblance to the ancient alchemistic aspirations of turning base metals into gold. That is not to say that there is no genuine scientific effort in studying financial markets, but distinguishing serious research from charlatanism (or even fraud) remains remarkably difficult.
  
  
• Breast Cancer Diagnostics with Bayesian Networks
  The Wisconsin Breast Cancer Database (WBCD) is a widely studied (and publicly available) data set from the field of breast cancer diagnostics. The creators of this database, Wolberg, Street, Heisey and Managasarian, made an important contribution with their research towards automating diagnostics with image processing and machine learning. Beyond the medical field, many statisticians and computer scientists have proposed a wide range of classification models based on WBCD. Such new methods have continuously raised the benchmark in terms of diagnostic performance. Our white paper now reevaluates the Wisconsin Breast Cancer Database within the framework of Bayesian networks, which, to our knowledge, has not been done before. We demonstrate how the BayesiaLab software can extremely quickly - and simply - create a Bayesian network model that is on par performance-wise with virtually all existing models that have been developed from WBCD over the last 15 years.
  
  
• Microarray analysis with Bayesian Networks
  In this study, we turn to the field of cancer classification by means of microarray analysis. Microarray analysis is a technique for gene expression profiling of cell samples. Expression profiles indicate which genes are currently active among thousands of genes.
  
  
• Modeling Vehicle Choice and Simulating Market Share with Bayesian Networks
  We present a new method and the associated workflow for estimating market shares of future products based exclusively on pre-introduction data, such as syndicated studies conducted prior to product launch. Our approach provides a highly practical, fast and economical alternative to conducting new primary research. With Bayesian networks as the framework, and by employing the BayesiaLab and Bayesia Market Simulator software packages, this approach helps market researchers and product planners to reliably perform market share simulations on their desktop computers, which would have been entirely inconceivable in the past.
  
  

 Download BayesiaLab 5.0 Trial

A free 30-day evaluation version of the latest release of BayesiaLab 5.0 Professional Edition is available for immediate download.

Click here to register and download BayesiaLab (Windows, OS X, Linux/Unix, 32/64-bit).

This will allow you to experiment with new the features of BayesiaLab 5.0, plus you can try out all the Bayesian network examples explained in our series of white papers.

 

 Upcoming Webinar

Thursday, December 15, 2011, 12 noon CST (GMT -06:00)

Probabilistic Structural Equation Models (PSEM), based on machine-learned Bayesian networks, provide an efficient alternative to traditional Structural Equation Models (SEM). With BayesiaLab 5.0, PSEMs can be created through a series of semi-automatic steps, which allow analysts to perform driver analysis extremely quickly, reducing research time from “months to minutes.” This Webinar will demonstrate a complete workflow for a typical application in the field of marketing science. Dr. Lionel Jouffe and Stefan Conrady will present several updates to the approach originally described in their white paper, Driver Analysis & Product Optimization. This will include an illustration of Direct Effects computed by means of Likelihood Matching.

Register for the Webinar here.

 

 Recorded Webinars

• Driver Analysis with Probabilistic Structural Equation Models
  Probabilistic Structural Equation Models (PSEM), based on machine-learned Bayesian networks, provide an efficient alternative to traditional Structural Equation Models (SEM). With BayesiaLab 5.0, PSEMs can be created through a series of semi-automatic clustering steps, which allow analysts to perform driver analysis extremely quickly, reducing research time from "months to minutes." This webinar will demonstrate a complete workflow for a typical application in the field of marketing science, namely driver analysis of purchase intent and product optimization based on consumer survey data.
  
  
• Knowledge discovery in the stock market with Bayesian Networks - Part 1
  We demonstrate how BayesiaLab can generate a Bayesian network describing the interactions in the stock market. More specifically, we will use BayesiaLab's Unsupervised, Supervised and Semi-Supervised Learning algorithms to learn the movements of all stocks in the S&P 500 index over a period of several years. The resulting networks provide intuitive representations of the dynamics within the S&P 500 index.
  
  
• Knowledge discovery in the stock market - Part 2
  We use again in that webinar the BayesiaLab's Unsupervised, Supervised and Semi-Supervised Learning algorithms for automatically building new Bayesian networks representing the relationships that hold between all stocks in the S&P 500 index over a period of six years.
  
  
• Data Clustering with Bayesian Networks and BayesiaLab 5.0
  We provide an introduction to data clustering with Bayesian networks. By using Bayesian networks as the analysis framework, BayesiaLab 5 is an extremely convenient and fast software tool for discovering structures and patterns within the data.
  
  
•  Bayesian Network 101
  This 30-minute Webinar will provide a brief overview of the basic concepts behind Bayesian networks. You will see how a few simple and intuitive ideas can form a marvelous framework for inference and reasoning. Among other fundamental topics, we will explain the connection between the factorization of a joint probability distribution and the structure of a Bayesian network. For illustration purposes we will use the well-known sprinkler example that is commonly presented in the literature.

 

 Bayesian Belief Networks Discussion Group

Last December we created a discussion group on LinkedIn dedicated to Bayesian Belief Networks with BayesiaLab. Today, this group has more more than 700 members and it hosts a wide range of interesting discussions and also features the latest announcements from Bayesia. You will also find Frequently Asked Questions (FAQ) in this group, with experts from BayesiaLab and users of the program exchanging the latest tips and tricks.

You can join our LinkedIn group clicking here.