Skills
Applied machine learning
Machine learning builds a rule from examples instead of coding it. Feature selection, training, honest evaluation, interpretability: four steps separate an experiment from a model a company is willing to decide on.
Learning from examples rather than rules
A model scoring brilliantly on its own training data mostly tells you it memorised them. The whole skill lies in how you evaluate: clean data splits, cross-validation, a metric chosen for the real cost of a business error. A false negative on a safety check and a false positive on a sales follow-up do not cost the same, and the model must reflect that trade-off. The teaching puts that requirement ahead of raw performance.
- Frame a business problem as supervised or unsupervised learning
- Engineer explanatory features from raw data
- Evaluate a model on data it has never seen
- Explain a prediction to a decision-maker who has to own it
Your level of command, stated plainly
You run a machine learning project from raw data to evaluated model, and justify every trade-off along the way. A skill is best stated by what it lets you carry out on your own, on completion.
- teaching blocks
- 4
- of taught instruction
- 650 h
- credits in total
- 65
- professional qualifications concerned
- 3
Catalogue rule: one credit is twenty-five hours of work, ten of them taught.
The taught hours, block by block
The catalogue rule applies throughout: one credit stands for twenty-five hours of work, ten of them taught. These rows give the exact translation.
| Teaching block | Programme | Taught hours | Year |
|---|---|---|---|
| Machine learning | Artificial Intelligence programme — Bac+5 level | 200 h | Year 1 |
| Modelling and interpretability | Data Science programme — Bac+5 level | 200 h | Year 2 |
| Mathematics for AI | Artificial Intelligence programme — Bac+5 level | 220 h | Year 1 |
| Understanding without coding | Applied AI for Business Certificate | 30 h | Year 1 |
What you can aim for with it
Each row opens a qualification page: blocks, volume, duration and entry requirements.
| Professional qualification | Target occupation | Level | Taught volume |
|---|---|---|---|
| Data and Applied AI Technician | data technician, AI solutions integrator | 5 · Senior technician and supervisor | 700 h |
| AI Project Manager | AI project manager, data lead | 6 · Manager and professional | 800 h |
| Chief Data and AI Officer | data director, chief data and AI officer | 7 · Expert and executive | 700 h |
Which programme takes you there
The same professional act can be reached by several routes. Compare volume, duration and amount.
| Programme | Level | Duration | Fees |
|---|---|---|---|
| Artificial Intelligence programme — Bac+5 level | Master's degree | 2 years | 7 800 000 GNF per year |
| Data Science programme — Bac+5 level | Master's degree | 2 years | 7 200 000 GNF per year |
| Applied AI for Business Certificate | Professional certification | 4 months | 3 200 000 GNF in total |
The questions we are asked
What level of mathematics is expected?
The programme opens with 220 taught hours of applied mathematics: linear algebra, probability, optimisation. They are taught for use, illustrated on the models studied next.
Do you work on real data?
Yes, on datasets from West African contexts, with the characteristics that come with them: modest volumes, missing values, heterogeneous reference systems. That is what prepares you for the field.
Can this skill be exercised without writing code?
The applied artificial intelligence certificate is for managers who frame and appraise projects without coding them. The long programme trains the person who builds them.
Explore next
- Deep learningDeep learning and language models at IHETC: 400 taught hours, from network architecture to document-grounded generation. Programmes and roles.
- MLOpsMLOps at IHETC: model deployment and monitoring, containers, versioning, drift and compute cost. Taught hours, programmes and roles concerned.
- DashboardsDashboards and reporting at IHETC: indicator selection, chart design, refresh and executive readability. Taught hours and target roles.
- LinuxLinux at IHETC: command line, permissions, services, logs and operations scripting. Taught in computer science, networks and cloud, and cybersecurity.
- TCP/IP networkingTCP/IP networking at IHETC: 240 taught hours of addressing, switching, routing, segmentation and continuity. Programmes concerned and target qualifications.
- Containers and orchestrationContainers, Kubernetes and Terraform at IHETC: virtualisation, orchestration, cloud architectures and deployment. Taught hours and target roles.