User Guide

• 1. Methods Taxonomy
  • 1.1. Aggregative Quantification
    • 1.1.1. Classify Count
    • 1.1.2. Mixture Models
    • 1.1.3. Threshold Methods
    • 1.1.4. Expectation Maximisation for Quantification
  • 1.2. Non Aggregative Quantification
  • 1.3. Meta Quantification
• 2. Label Scheme
  • 2.1. Binary problems
  • 2.2. Multiclass Problems
• 3. Building a Quantifier
  • 3.1. Building a Quantifier
  • 3.2. Building an Aggregative Quantifier
  • 3.3. Building a Non-Aggregative Quantifier
• 4. Non Parameters Scenarios
• 5. Model Selection and Evaluation
  • 5.1. Tunning the hyper-parameter of a quantifier
  • 5.2. Metrics and scoring: quantifing the performance of a quantifier
• 6. Experiment Management
  • 6.1. Protocol
  • 6.2. Artificial Prevalence protocol with general usage
  • 6.3. Artificial Prevalence protocol with selected usage
  • 6.4. Building a Protocol