The aim of the unit is to study different probability models and specific algorithmic methods for data learning and to apply them with R software
On completion of the unit, the student will be capable of: | Classification level | Priority |
---|---|---|
Understanding the different statistical models of supervised and non-supervised learning and the validation methods of these models | 2. Understand | Essential |
Knowing how to recognize the nature of the set problem and to apply the appropriate models | 3. Apply | Essential |
Analysing and of criticising obtained results | 4. Analyse | Essential |
Producing a professional report, summarising the obtained results | 6. Assess | Important |
Percentage ratio of individual assessment | Percentage ratio of group assessment | ||||
---|---|---|---|---|---|
Written exam: | 60 | % | Project submission: | 0 | % |
Individual oral exam: | 0 | % | Group presentation: | 0 | % |
Individual presentation: | 0 | % | Group practical exercise: | 0 | % |
Individual practical exercise: | 40 | % | Group report: | 0 | % |
Individual report: | 0 | % | |||
Other(s): 0 % |
Type of teaching activity | Content, sequencing and organisation |
---|---|
Course | Generalities: bias- variance dilemma, model validation |
Course | Non-supervised learning (in addition to the Exploratory statistics and mathematical tools unit): association rules In addition to the theoretical course a practical case study will be undertaken. |
Course | Supervised learning of approximation: linear methods (regressions), neurone networks, trees, etc. Application to a case study. |
Course | Supervised learning of classification: linear methods (discriminatory analysis, logistics regression), closest neighbour methods, support vector machines (SVM). Application to a case study |