|
Teacher
|
cotroneo rossana
(syllabus)
1. Data: structure and manipulation
• Statistical data
• The IT data
• Data manipulation: random and systematic errors; missing data; outlier; logical inconsistencies data; anomalous data; imputation technique for missing data.
2. The syntheses of the data, their transformations
3. multivariate distributions:
• multinormal distribution
• multinomial distribution
4. Unsupervised classification (cluster analysis)
- Non-hierarchical methods: K-means, Pam and fuzzy K-madoidi
- variables Selection, best selection of the number of clusters,
- The aggregative hierarchical methods
5. Data reduction methodologies:
• principal component analysis (PCA);
• factor analysis
• correspondence analysis and multiple correspondence analysis
6. Unsupervised classification (cluster techniques)
7. Supervised classification (cluster techniques)
o Linear regression;
o Generalized linear model
o Logistic regression;
o Classification trees
o Support vector machine
o Random forest
o Gradient boosting
o Neural networks
• Traning, Evaluating and Testing model
(reference books)
- slide delle lezioni che saranno caricate sulla piattaforma Moodle
|
