What is the point of our IA and software testing training program ?
The classical approach to software testing is based on a scripted approach.
The creation of these test plans / expected application behaviour requires time. The problem of executing tests, which are usually repetitive, is addressed by test automation tools.
However, tests can still represent a bottleneck in the execution of a project in an Agile environment.
In recent years, the use of “deep learning” and artificial intelligence (AI) has developed in order to provide business solutions based on data analysis to predict behaviour or solve specific technical problems (shape recognition, autonomous driving, business process automation, etc.).
The two issues that arise are therefore:
- How to test an application that uses AI / Deep learning?
- How to use AI to help teams test an application differently and save time?
It is to answer these questions that ALL4TEST – in collaboration with the A4Q association – offers a certified training on AI and testing.
Following this training, we also offer, as an option, a coaching / POC tools to help you implement these technologies on your IT projects. Contact us for more information!
A. The content of the IA and software testing training program
1. The outline of the IA and testing training program
Initially, the key aspects of AI such as its history, symbolic AI (i.e. visible to humans) or its limits; are covered in the first of the three main chapters of the Syllabus.
Then, once the prerequisites in terms of AI language are understood, it is time to move on to the testing of AI systems. All participants will then learn about the different measures and strategies to be implemented to test these systems, as well as the issues related to these tests.
Finally, these teachings are completed by modules on:
- the use of AI for test support
- the method of applying AI to test tasks and quality management.
The A4Q AI and Software Testing Foundation certification provides a professional level of knowledge and understanding of the use of Artificial Intelligence in software testing and the use of AI in applications that are based on this technology.
2. The objectives of this training program
- Know the key aspects of Artificial Intelligence
- Know how to conduct tests of systems that incorporate AI
- Be able to use artificial intelligence to support testing
3. People who might be interested in this IA and software testing training program
- Software engineers and testers
- Designers in general and usability designers in particular
- Anyone who is interested or involved in AI-enabled software testing and wants to maximize their understanding
- Business analysts and managers looking to understand how artificial intelligence could add value to the business
B. Details about the A4Q AI and Software Testing certification exam
1. How the exam is administered
This exam is composed by of 40 multiple choice questions, with one point for each correct answer. The requirement for passing this exam is a score of at least 65% or at least 26 questions with a correct answer. The time limit for taking the exam is one hour. Additional time may be granted to candidates whose native language is not the language of the exam (25% or 15 minutes).
2. Recommendations to follow
GASQ’s recommendation for participants is to complete a training program with an accredited provider before taking the exam.
In order to organize your exam, ALL4TEST is at your disposal to help you have all the necessary elements to succeed in your IA and software testing training and your A4Q certification.
Click here to see the Syllabus (English version).
C. Below you will find the summary of the Syllabus for the AI and Software Testing Training
Introduction
Purpose of this syllabus
Examinable learning objectives and cognitive levels of knowledge
The AI and Software Testing Foundation exam
Accreditation
Level of detail
Organization of this syllabus
Business outcomes
Acronyms
1.0 Key Aspects of Artificial Intelligence
Keywords
Learning objectives for key aspects of artificial intelligence
1.1 What are human intelligence and artificial intelligence?
Types of intelligence
Turing test
1.2 History of AI
Main periods in the history of AI
Difference between symbolic and sub-symbolic AI
1.3 Symbolic AI
Mathematical logic and inference
Knowledge-based systems
Constraint-based problem solving systems
1.4 Sub-symbolic AI
Types of learning
Examples of applications of different types of learning
Machine learning algorithms
Machine learning metrics
1.5. Some ML algorithms in more detail
Bayesian belief networks
Naïve Bayes Classifier
Support vector machine algorithm
K-means algorithm
Artificial Neural Networks: Perceptron Learning Algorithm
1.6. Applications and Limitations of AI
Machine learning activities
Possible biases in AI systems
Ethical issues in AI systems
2.0 Testing of artificial intelligence systems
Key words
Learning objectives for testing artificial intelligence systems
2.1 General issues in testing AI systems
Software written to compute an output for which the correct answer is not known
Real-world inputs
Self-optimization
Expert systems
Perception of intelligence
Model optimization
Quality characteristics of AI systems
Bias/variance trade-off and no free lunch theorem
Drift
Ethical considerations
Automation Bias
Adversarial actors
2.2 Training and testing of machine learning models
2.3 AI test environments
2.4 Test strategies for AI-based systems
Acceptance criteria
Functional testing
External and internal validity
Metamorphic testing
A/B testing
Evaluation of real-world results
Expert Panels
Test Levels
Component Testing
System Integration Testing
System Testing
User acceptance testing
2.5 Test metrics for AI-based systems
Confusion Matrix
Statistical significance
3.0 Using AI to support testing
Keywords
Learning objectives for using AI to support testing
3.1 AI in Testing
The Oracle Problem
The oracles of testing
Testing versus test automation
3.2. Application of AI to testing tasks and quality management
Tasks to which AI can be applied
Tasks to which AI cannot be applied
Using AI for Test Data Generation
Use of AI for bug triage
Use of AI for risk prediction and failure estimation
3.3 AI in component-level test automation
AI in component-level test generation
AI in system-level test generation
3.4 AI in integration or system level test automation
Monkey Testing Versus Fuzz Testing
AI for system level test generation
AI for test selection and prioritization
AI for object identification and identifier selection
AI for visual test automation
3.5 Support for AI-based testing tools
Relevant metrics in an AI-based testing approach
Evaluate the tool vendor’s claims
System configuration
Return on Investment (ROI)
Effects on existing processes
Test case sensitivity
Test case explosion
Maintainability
Severity of defects found
Summary
Ask for advice to our experts !
