Choosing the Right ML Approach for Your Business Case with ISO/IEC 25053:2022

Course details

In this course, instructor Lyron Andrews shows you how to describe the system components of machine learning (ML) and their function in the AI ecosystem. This includes identifying general, supervised, unsupervised, and semi- or self-supervised learning, along with enumerating the steps related to the ML pipeline. Along the way, gather insights for evaluating the need for anomaly detection and dimensionality reduction related to training models. Lyron outlines the steps of the machine learning pipeline, from data acquisition and preparation to modeling, verification, and validation. By the end of this course, you'll be prepared to choose the right ML approach to meet the unique needs of your business case.

Learning objectives
Analyze a dataset to identify potential patterns and anomalies that could impact your machine learning model.
Create a comprehensive machine learning pipeline that incorporates all necessary steps from data acquisition to model validation.
Evaluate the performance of different supervised learning algorithms on your specific business problem to determine the most effective approach.
Apply dimensionality reduction techniques to improve the efficiency and accuracy of your machine learning models when dealing with high-dimensional data.
Differentiate between the various types of machine learning approaches and explain when each is most appropriate to use in real-world scenarios.

Concepts

0. Introduction

01 - Why the need for an ML process approach

1. Terms, Definitions, and Overview

02 - Overview of clauses, terms, and definitions (Clauses 3-5)

2. Machine Learning Systems

03 - Task and general (Clause 6-6.2)
04 - Task details (Clauses 6.2.2-6.2.7)
05 - ML model (Clause 6.3)
06 - ML data (Clause 6.4)
07 - Tools and data preparation (Clauses 6.5-6.5.2)
08 - Support vector machines ML (Clause 6.5.3.5)
09 - Bayesian ML (Clause 6.5.3.3)
10 - Decision tree ML (Clause 6.5.3.6)

3. Categories of ML Neural Networks

11 - General and FFNN ML (Clauses 6.5.3.2-6.5.3.2.2)
12 - RNN and LSTM ML (Clauses 6.5.3.2.3-6.5.3.2.3.2)
13 - CNN and CapNet ML(Clauses 6.5.3.2.4, 6.5.3.2.7)
14 - DBM, structured perceptron, and GAN ML (Clause 6)
15 - ML optimization methods (Clauses 6.5.4-6.5.4.8)
16 - ML evaluation metrics (Clauses 6.5.5-6.5.5.8)

4. Machine Learning Approaches

17 - General, supervised, and unsupervised (Clauses 7-7.3)
18 - Semi- and self-supervised (Clauses 7.4-7.5)
19 - Reinforcement and transfer (Clauses 7.6-7.7)

5. Machine Learning Pipeline

20 - Data acquisition and preparation (Clauses 8-8.3)
21 - Data preparation details (Clause 8.3)
22 - Modeling, verification, and validation (Clauses 8.5-8.6)
23 - Model deployment and operation (Clauses 8.6-8.7)
24 - Machine learning pipeline example (Clause 8.8)

Conclusion

25 - Continue developing your ML business cases

Choosing the Right ML Approach for Your Business Case with ISO/IEC 25053:2022

Authors

Lyron Andrews