San Francisco, California
London, United Kingdom
Many BayesiaLab users have been asking for a course covering BayesiaLab's advanced features that are outside the scope of the 3-day introductory course. For the first time, and in conjunction with the BayesiaLab User Conference in Orlando, we are offering such an advanced program. The instructor, Dr. Lionel Jouffe, will be focusing on many of the recently introduced or enhanced features of BayesiaLab, such as knowledge elicitation and optimization. The advanced course is a 2-day program, scheduled to run from 9 to 5 on both days.
Manual Modeling with BEKEE Description of the workflow Exercise with the group of trainees Probability Table Analysis on Target State Parameter Sensitivity Analysis Principle Exercise with a Fire Detection System Influence Diagrams Principle Example: Oil Rig Exercises: Monty Hall Problem, Investment Policy Dynamic Bayesian Networks Unfolded (Hidden Markov Chain example) Compact Form Temporal Simulation Exact and Approximate Inference Exercise with the Battery Example Network Temporalization Example with Box & Jenkins Forecast Bayesian Updating Principle Unfolded Form Compact Form Exercise with the Horse Doping Example Data Import Discretization Manual Automatic Rule of thumb for choosing the best method Aggregation Missing Values Processing Filtering Replacement Static Imputation Dynamic Imputation Structural EM Entropy Based Static Imputation Entropy Based Dynamic Imputation Filtered States Imputation Standard (law or Maximum a posteriori) Entropy based (law of MAP) Most Probable Explanation Exercise Creation of a network modeling MCAR, MAR and NMAR missing values, and Filtered States Data generation and learning Synthesis of New Variables Manual creation Binarization KMeans and Bayesian Clustering on a subset of variables Binary Clustering Semi-Supervised Data Clustering Hierarchical Clustering Exercises Variable Clustering Cross-validation Comparison of solutions Fine Tuning of Learning Virtual Number of States Structural Coefficient Local Structural Coefficient Stratification Variable Selection with Cross-Validation Smooth Probability Estimation Analysis Evidence Analysis Likelihood Analysis Exercise for Mutli-Dimensional Outliers Detections Straw Models Analysis with the Less Probable Evidence Supervised Learning Maximum Probable Explanation Path Analysis Information Analysis Optimization Target Optimization Contributions Direct Effect Contributions Type I and Type II Contribution Analysis based on Counterfactuals Exercise on a Marketing Mix Example Negative and Disjunctive Inference Mapping Exercise Evidence Instantiation and Evidence Data Weighting Principle Exercise Design of Experiments Target Global
Terms & Conditions
- The tuition fee for the 2-day course for commercial participants is US$ 1,995.
- Members of government agencies, the military and non-profit organizations are eligible for a reduced tuition fee of US$ 1,495 (25% discount).*
- Student and faculty of accredited academic institutions are eligible for a reduced tuition fee of US$ 995 (50% discount).*
- The course fee includes a Conference Pass for the BayesiaLab User Conference on October, 24, at the same venue as the course.
- A 60-day license to the full version of BayesiaLab Professional 5.1 will be provided to all participants for installation on their computers prior to the event.
- Participants will be required to bring their own WiFi-enabled computer/laptop to the seminar (Windows XP/Vista/7/8 or Mac OS X).
- The course fee includes all training materials, beverages, lunch and snacks during the training.
- Accommodation in Orlando is at the participants' own expense, although negotiated rates are available at the seminar venue.
*Proof of affiliation will be required upon registration. If you are not sure about your eligibility for reduced tuition fees, please email us at firstname.lastname@example.org.
When & Where
Bayesia USA is the North American sales and consulting organization for France-based Bayesia S.A.S. Their mission is to promote Bayesian networks as a new framework for knowledge discovery and reasoning within complex domains.
Founded by two professors in the field of artificial intelligence in 2001 and headquartered in northwestern France, Bayesia S.A.S. is the world's leading developer of research software based on the Bayesian network paradigm. Their principal product, BayesiaLab, is the only software platform that can perform unsupervised structural learning for knowledge discovery.
Today, BayesiaLab is being used by researchers in major organizations around the world, including P&G, Unilever, BBDO, GroupM, GfK, TNS, Ipsos, Mu Sigma, Booz Allen Hamilton, InterContinental Hotels Group, Dell and NASA's Jet Propulsion Laboratory among many others.