fr
en
Understanding AI algorithms and logic
Contributing to an AI project requires the ability to distinguish real capabilities from hype. Connect data, models and use cases to understand the choices involved and their limitations. Build a foundation for more precise discussions with technical teams and better-supported decisions.
- Duration
- 1 day 7 hours
- Code
- IA032FR Code
Presentation
Artificial intelligence is often seen as a complex black box. Yet understanding its internal logic is essential for working effectively with technical teams and making informed decisions. This one-day course explains the fundamentals of AI algorithms in accessible terms, without requiring a technical background.
The program guides you step by step through the mechanisms of machine learning. Discover how a machine learns, distinguishes and predicts by exploring the main algorithm families: supervised, unsupervised and reinforcement learning. The teaching approach prioritises clarity and practical illustrations to demystify technical jargon.
Through engaging workshops and simulations, including role-play and paper-based modelling, you will experience how a model is built, from data preparation to critical interpretation of the results. Leave with the knowledge needed to discuss ideas with data experts and assess the relevance of an algorithmic solution.
Objectives
By the end of this course, you will be able to:
- define an algorithm and its specific role within artificial intelligence;
- distinguish the main learning families—supervised, unsupervised and reinforcement learning—according to the requirement;
- explain the logic behind classification, regression and clustering tasks;
- identify the key stages in building and training an AI model;
- interpret algorithm results while identifying potential limitations and biases.
Program
Module 1: discovering the algorithmic foundations of AI
- Defining an algorithm and distinguishing conventional programming from AI.
- The complete lifecycle of an artificial intelligence algorithm.
Hands-on exercises
- Simulate a manual sorting algorithm through a role-playing exercise.
Module 2: exploring machine learning approaches
- The principle of supervised learning: classification and regression techniques.
- How unsupervised learning and clustering work.
- The logic of reinforcement learning for decision-making.
Hands-on exercises
- Identify the most suitable learning approach for different business use cases.
Module 3: building a simple model
- Selecting an appropriate algorithm for the identified requirement.
- The crucial importance of data preparation and quality.
- Model training and validation stages.
Hands-on exercises
- Simulate the construction of a prediction model using a paper-based exercise or a simplified tool.
Module 4: interpreting results and limitations
- Reading performance indicators, including precision, the confusion matrix and error.
- Detecting algorithmic bias and overfitting.
- The direct impact of data quality on decision reliability.
Hands-on exercises
- Critically analyse an algorithm's output to identify weaknesses.
Audience
This course is intended for business professionals and decision-makers, including:
- product owners and project managers seeking to manage data products more effectively;
- business managers who want to understand the mechanisms behind the solutions they use;
- AI beginners seeking a solid foundation in how algorithms work;
- anyone involved directly or indirectly in AI or data projects.
Prerequisites
The following background is useful:
- Basic knowledge: general project management or IT knowledge helps put the learning into context.
Teaching and assessment methods
- Initial skills assessment
- Training materials provided to participants
- Continuous assessment throughout the course
- End-of-course feedback questionnaire
- Combination of theory and practical application
- Attendance records
- Post-course follow-up evaluation
- Practical exercises
- Case study
Course highlights
- Accessible explanations: explore complex concepts explained simply, without mathematical or technical barriers.
- Learning through play: understand algorithmic logic through manual simulations and role-playing exercises.
- Critical perspective: learn not only how models work but, more importantly, how to assess their reliability and risks.
- Short format: acquire essential algorithmic literacy in one intensive day.
Dates and sessions
Choose the date and delivery format that suit you.
No upcoming sessions are currently available.