Geographic Optimization with Bayesian Networks

Had not seen this kind of optimization before with Bayesian Networks.  Webinar leads you through the process, largely non-technical.

 ... By Stefan Conrady, 
Managing Partner at Bayesia USA & Singapore: Bayesian Networks for Research, Analytics, and Reasoning .... 

Geographic Optimization with Bayesian Networks and BayesiaLab
You may not know that you can use BayesiaLab for geographic optimization. Today's webinar explained how you can find an optimal location for a distribution hub that needs to connect thousands of geographically dispersed suppliers and customers. Bayesian networks and BayesiaLab make this type of optimization remarkably quick and easy. ...  "

More Bayesian Networks

Video, another example of using Bayesian networks.

Dr. Alta de Waal's presentation at the 5th Annual BayesiaLab Conference in Paris.

Spatially Discrete Probability Maps for Anti-Poaching Efforts

Via Stefan Conrady
Managing Partner at Bayesia USA & Singapore: Bayesian Networks  ...

Bayesian Networks for Marketing

Some good and fairly straightforward examples of the technology in marketing.

Bayesia Presents:
Benoit Hubert: From Marketing Science to Artificial Intelligence with Bayesian Networks

Benoit Hubert (GfK) presents at the 5th Annual BayesiaLab Conference in Paris

Free Bayesian Networks Research eBook

New Book Release:
from Bayesian Networks & BayesiaLab

A Practical Introduction for Researchers By Stefan Conrady and Lionel Jouffe,  382 pages, 433 illustrations

Our long-awaited book is now available for free download. It was released just in time for the 2015 BayesiaLab Conference in Fairfax, Virginia.

Link at: http://www.bayesialab.com/book  Requires registration

A hardcopy version of the book is available for ordering via Amazon. .. " 

Delivering Bayesian Network Simulation Models

Bayesian network simulation was a favorite analytical method at our enterprise, they still maintain an active presence.   Now in a new release BayesiaLab has a means to Web deliver Bayesian network models. Nice idea.   Any simulation has to ultimately be delivered to the business decision makers to validate its assumptions and understand its potential value.  An example of getting closer to automated data science.  Easy to trial.

New Release: BayesiaLab 5.4

The most important new feature of BayesiaLab 5.4 is the BayesiaLab WebSimulator. It allows you to publish interactive models via the web, without having to install any additional software for your audience. Any Bayesian network model you build with BayesiaLab can now be shared privately* with your clients, or publicly with the wider world. 

Once you publish your network via the BayesiaLab server, your audience can simulate scenarios and intuitively explore the properties and dynamics of your Bayesian network model. ... " 

Full Announcement page.

Bayesian Networks as Expert Systems

Modeling causal reasoning is a favorite topic for me. Bayesia has posted another easy to understand demo of how a Bayesian network can be used to model a system of expertise.  Nicely done.  My former enterprise uses this methodology.  If how this fits into other expertise capture work is still confusing to you, this is a good place to start:

" ... Using a Bayesian Network as an Expert System ... 
The well-known "Visit Asia" example in this Quick Start Demo dates back to a seminal paper by Lauritzen and Spiegelhalter (1988). It describes how a Bayesian network can be used as an expert system for medical diagnosis. Visit Asia has been widely used in the literature to illustrate observational and causal inference in Bayesian networks. ... "

Knowledge Discovery with Unsupervised Learning

Fascinating example using Bayesia's graphical methods.   I like how the visualization leads to useful  implications of the discovery.  Easy to follow example with stock market data.  Links to a complete tutorial.   Why Stocks?  They motivate:

" .... This area can perhaps serve as a very practical proof of the powerful properties of Bayesian networks, as we can quickly compare machine-learned findings with our own understanding of market dynamics. For instance, the prevailing opinions among investors regarding the relationships between major stocks should be reflected in any structure that is to be discovered by our algorithms.

More specifically, we will utilize the unsupervised and supervised learning algorithms of the BayesiaLab software package to automatically generate Bayesian networks from daily stock returns over a six-year period. We will examine 459 stocks from the S&P 500 index, for which observations are available over the entire timeframe. We selected the S&P 500 as the basis for our study, as the companies listed on this index are presumably among the best-known corporations worldwide, so even a casual observer should be able to critically review the machine-learned findings. In other words, we are trying to machine-learn the obvious, as any mistakes in this process would automatically become self-evident. Quite often experts’ reaction to such machine-learned findings is, “well, we already knew that.” That is the very point we want to make, as machine-learning can — within seconds — catch up with human expertise accumulated over years, and then rapidly expand beyond what is already known.expertise accumulated over years, and then rapidly expand beyond what is already known. .... " 

Judea Pearl Keynote Address

Keynote Lecture at the 2014 BayesiaLab User Conference

September 23, 2014, Los Angeles, California

From Bayesian Networks to Causal and Counterfactual Reasoning  a talk by Judea Pearl.

"... The development of Bayesian Networks, so people tell me, marked a turning point in the way uncertainty is handled in computer systems. For me, this development was a stepping stone towards a more profound transition, from reasoning about beliefs to reasoning about causal and counterfactual relationships. In this talk, I will survey the milestones of this journey, and  summarize the practical and conceptual problems that we can solve today and could not address two decades ago. ... "

Registration required.

Autonomy Accounting

Remember visiting HP Autonomy in the UK in the 90s.  Was impressed what they showed us. Analytics being used with AI and Bayesian methods.  First time I had seen these approaches being used with real corporate data. Proposed it for use, but it never happened, in part due to the financial flap.  Now was this due to mere accounting differences and not fraud?  And all much ado about nothing?  Too bad if so, I liked what I saw then.

Big Data Must Go Causal

Correspondent Stefan Conrady of Bayesialab writes about the need for big data to go causal.   I have been reading much about the topic recently  See the Causation tag below.  He writes " ...  we are substantially expanding our educational program in 2015, emphasizing the critical importance of causal inference in the world of analytics. Unfortunately, many data scientists remain unconcerned about the critical distinction between statistical and causal inference. There is still little awareness of the requirements for proper causal identification and estimation. ... " 

More on this and their educational program.

Causality for Policy Assessment and Impact Analysis

From BayesiaLab.     Nicely done live recorded lecture.  To get the full extensive tutorial you need to register.  The video is open without registration.

New Video Lecture: Causality for Policy Assessment and Impact Analysis
Presenters: Stefan Conrady and Dr. Lionel Jouffe
Recorded on November 18, 2014, at George Mason University Arlington Campus.
Runtime: 01:52:24

The objective of this presentation is to provide a practical framework for causal effect estimation with non-experimental data. We will present a range of methods, including Directed Acyclic Graphs and Bayesian networks, which can help distinguish causation from association when working with data from observational studies. The presentation revolves around a seemingly trivial example, Simpson’s Paradox, which turns out to be rather tricky to interpret in practice.

This talk is a "live" version of a recent tutorial, Causality for Policy Assessment and Impact Analysis - Directed Acyclic Graphs and Bayesian Networks for Causal Identification and Estimation. .... " 

Thinking Causality in Science and Statistics

I have been looking back to understand how AI has changed since the 90s, when we worked with rule based expert systems.  One book that addresses some of the changes is Judea Pearl's:  Causality: Models, Reasoning and Inference.  Now over a decade old, it contains some interesting gems. Dealing with the mixing of knowledge in diagrams and equations, and developing approaches that have evolved to now commonly used Bayesian Networks.  More on his site.

There is also a copy of a lecture that Pearl gave at the time:  The Art and Science of Cause and Effect, originally an epilogue in the book.   Now available free at the link. Deals with the interesting concept of Causality, which is remarkably complex.  The idea is essential in working engineered systems, avoided in the physical sciences, and most always warned against in statistics.  The article examines why, and poses some remedies.  I disagree with some of his early historical views, causation was not discovered at the time of Galileo, but the lecture is still an excellent read.

Consider also how Big Data methods have backed off the need for strict causation requirements.

Big Data in Sports

In Linkedin: How Germany Beat Brazil. And how Big Data Helped.  Via my colleague : Ali Rebaie: Big Data Analyst and Consultant , Top Big Data Influencer, Public Speaker and Trainer.   See also SAP's alleged involvement with World Cup big data.  Also how Bayesian networks have been used in this space.

Predicting World Cup Outcome with Bayesian Networks

A nice example of how a Bayesian network can be used to do predictive analytics.  Much more here.  " ... Designed around a Causal Bayesian Network, and using the BayesiaLab's API, the BWCP computes each team's qualification probabilities for proceeding from the Group Stage to the Knockout Stage in the 2014 FIFA World Cup. ... ".    At the link you can give it a try.

Using Bayesian Methods to Find the Malaysian Airplane

Via the SAS Blog.  Good example of predictive analytics with a not generally well known technique for Big Data practitioners.  Set in a real world tactical problem.   Good practical details.  Have written about this method many times in this blog.

How Bayesian analysis might help find the missing Malaysian airplane .... 

At the time this blog entry was written, there still appears to be little to no signs of locating the missing Malaysian flight MH370. The area of search, although already narrowed down from the size of the United States at one point to the size of Poland, is still vast and presents great challenges to all participating nations. Everything we’ve seen in the news so far have been leads that turn out to be nothing but dead ends. .... " 

Optimizing Customer Loyalty Using Bayesian Networks

In the Latest Bayesia Newsletter.   For optimization of Loyalty.  An approach I had not seen in this area before.   " ... This tutorial illustrates an innovative market research workflow for deriving marketing and product planning priorities from auto buyer surveys. In this study, we utilize the Strategic Vision New Vehicle Experience Survey, which includes, among many other items, customers’ satisfaction ratings with regard to over 100 individual product attributes ... " 

Topic Modeling in Machine Learning

I see that David Blei has been given the ACM-InfoSys award.  In reading about his work in textual analysis I saw this was related to work we had done in 'content analysis' years ago, analyzing text from consumer comments.   But making it far more sophisticated and useful. This is an extension of what is called topic modeling now.  Here is a non technical piece on topic modeling.   Blei et al wrote a paper on their extensions, which use Bayesian methods.   Here is a technical description of  that.  And advanced topic models in R.  All this is worth understanding for the state of this art today.  More in this blog on text analytics.

" ... David Blei is the recipient of the 2013 ACM-Infosys Foundation Award in the Computing Sciences. He initiated an approach to analyzing large collections of data using innovative statistical methods, known as "topic modeling," that make it possible to organize and summarize digital archives at a scale that would be impossible by human annotation.  His work is scalable to collections of billions of documents and has inspired new research programs across multiple disciplines, with applications for email archives, natural language processing, information retrieval, computational biology, social networks, and robotics as well as computational social sciences and digital humanities. ... " 

Faculty page at Princeton.

I see that Blei also has an excellent relatively non-technical introduction to Topic Modeling and pointer to additional resources that I am exploring.

Loyalty Optimization with Bayesian Networks

Another interesting example of the use of Bayesian networks for data mining. via Stefan Conrady of Bayesia.   Nice understandable draft tutorial.  " ... Identifying Priorities for Maximizing Repurchase Intent ... This tutorial illustrates an innovative market research workflow for deriving marketing and product planning priorities from auto buyer surveys. In this study, we utilize the Strategic Vision New Vehicle Experience Survey, which includes, among many other items, customers’ satisfaction ratings with regard to over 100 individual product attributes. ... " 

On Applied Concept Mapping

Some time ago I mentioned the book Applied Concept Mapping, by Brian Moon, et al  and also posts about concept mapping in general.   Also about how it can be implemented on a tablet.  And here is a concept map about the book itself.   Currently looking at how to link this idea to maps which describe linkages quantitative relationships, like Bayesian nets.  More details to follow as this evolves.

Modeling Vehicle Choice and Simulating Market Share with Bayesian Networks

Via Stefan Conrady of Bayesia: Modeling Vehicle Choice and Simulating Market Share with Bayesian Networks    Useful example of the technical application of Bayesian. Downloadable overview whitepaper.   See also their latest newsletter.

" ... 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. ... "