Child pages
  • Advanced Course

Contents

Knowledge Modeling, Causal Analysis and Data Mining with Bayesian Networks

Training Overview

  • Teaching objectives: Advanced knowledge modeling, machine learning and analysis methods with Bayesian networks and BayesiaLab
  • Length: 3 days
  • Required Level: Participants in the Advanced Course are required to have completed the Introductory Course.
  • Teaching methods: Tutorials with practical exercises using BayesiaLab plus plenty of one-on-one coaching
  • Trainer: Dr. Lionel Jouffe, CEO, Bayesia SAS.
  • Training materials: A printed tutorial (approx. 200 slides), plus a memory stick containing numerous exercises and white papers
  • Bayesian network Software: Bayesia provides all trainees with an unrestricted 60-day license of BayesiaLab Professional Edition, so they can participate in all exercises on their own laptops
  • Cost: 2100 Euros/Trainee. Discounts are available for groups of trainees of the same company. A special academic discount of 50% is also available for students and teachers of accredited educational institutions. 

Registration

The registration is complete upon payment of the fee by Bank Transfer, or Credit Cards. Visit the BayesiaLab Store to get the prices corresponding to the type of your organization and number of seats your are interested in.

Training Program

Day 1

  • Bayesia Expert Knowledge Elicitation Environment
    • Interactive
    • Batch
    • Analysis of Expert Assessments
    • Sensitivity Analysis of the Votes
  • Probability Table Analysis
  • Parameter Sensitivity Analysis


  • Context
  • Utility Nodes
  • Decision Nodes


  • Context
  • Hidden Markov Chain
  • Unfolded Dynamic Bayesian Networks
  • Compact Dynamic Bayesian Networks
  • Network Temporalization
  • Temporal Forecast

Day 2


  • Context
  • Unrolled Networks
  • Compact Networks


  • Context
  • Impact of Discretization
  • Discretization Methods
    • Expertise
    • Equal Distance
    • Equal Frequency
    • KMeans
    • Density Approximation
    • Tree
  • Aggregation Methods
    • Expertise
    • Semi-Automatic
    • Tree


  • Context
  • Types of Missingness
  • Types of Methods
    • Static
      • Filtering
      • A priori Replacement
      • Entropy Based and Standard Static Imputation
    • Dynamic
      • Dynamic Imputation
      • Structural Expectation-Maximization
      • Entropy Based Dynamic Imputation
      • Approximate Dynamic Imputation with Static Imputation


  • Context
  • Classical Solutions
  • BayesiaLab’s Solution

Day 3


  • Context
  • Manual Synthesis
  • Binarization
  • Clustering
    • Binary Clustering
    • KMeans
    • Bayesian Clustering
    • Hierarchical Bayesian Clustering


  • Context
  • Structural Coefficient
    • Global
    • Local
  • Virtual Number of States
  • Stratification
  • Smooth Probability Estimation


  • Context
  • Joint Probability
  • Global Consistency/Conflict Measure
  • Local Consistency
  • Bayes Factor
  • Log-Likelihood
  • Most Probable Explanation
  • Path Analysis


  • Context
  • Objective Function
    • Probability
    • Mean
    • Maximization - Minimization
    • Target Value
    • Resources
    • Joint Probability
  • Search Methods
    • Hard Evidence
    • Soft Evidence
    • Direct Effects
  • Genetic Algorithm


  • Context
  • Direct Effects
  • Type I Contribution
  • Type II Contribution
  • Base Mean
  • Normalization
  • Stacked Curves
  • Synergies


  • Disjunctive Evidence
  • Negative Evidence
  • Evidence Instantiation
  • Evidence Data Weighting


  • Context
  • Solution
    • Targeted
    • Global