Recommendation ITU-T P.1402 (07/2022) Guidance for the development of machine-learning-based solutions for QoS/QoE prediction and network performance management in telecommunication scenarios
Summary
History
FOREWORD
Table of Contents
1 Scope
2 References
3 Definitions
     3.1 Terms defined elsewhere
4 Abbreviations and acronyms
5 Conventions
6 Application of machine learning and artificial intelligence in QoS/QoE related measurements and network performance's management
     6.1 General best practices for applied ML
     6.2 Brief overview on machine learning
          6.2.1 Most common learning techniques
          6.2.2 Most commonly used ML algorithms
     6.3 Use cases for applied machine learning
     6.4 Minimum requirements for ML-based solutions
     6.5 Overview on ML optimization process
          6.5.1 Learning curves
               6.5.1.1 Supervised learning
               6.5.1.2 Deep learning
          6.5.2 Inference test for ML overfitting/underfitting
     6.6 Guidance on the evaluation and validation of ML-based solutions for QoS/QoE predictors and network performance management
          6.6.1 ML-based network management models
          6.6.2 ML-based QoS/QoE predictors
     6.7 Conclusions
Annex A  ML overfitting/underfitting test for decision tress
     A.1 Learning curves
     A.2 ML model's parameters test for over/underfitting
          A.2.1 Number of trees
          A.2.2 Number of features
          A.2.3 Bootstrap
          A.2.4 Bootstrap features
          A.2.5 Minimum leaf size
Bibliography
<\pre>