Introduction to Machine Learning

Prova orale

Module 1 (20hrs) Foundations of supervised and unsupervised classification and neural networks (Prof. Flandoli and Dr. De Palma)

1)      Introduction to machine learning: problem class, assumptions, evaluation criteria, model type, parameter fitting, algorithm

2)      (unsupervised learning) Principal Component Analysis (foundations and implementation by software R)

3)      (u. l.) complements on PCA, few ideas about clustering.

4)      (supervised learning) multiple linear regression (foundations and implementation by R)

5)      (s. l.) logistic regression and classification by regression

6)      (s. l.) classification by linear perceptron.

7)      Feature representation, gradient descent, stochastic gradient descent

8)      Neural networks: basic element, networks, activation function, back-propagation, training

9)      Neural networks: loss function, optimization, regularization

10)   Convolutional neural networks: filters, max pooling, architectures

Module 2 (20hrs) Application of Machine Learning algorithms in Bioinformatics and Life Sciences  (Dr. Raimondi)

1)      protein structure/function prediction using machine learning

2)      protein structure prediction using deep learning and co-evolution

3)      application of graph neural network for the prediction of protein interaction networks

4)      deep learning applications to genomics :DNA motif discovery(with hands-on in Python)

5)      deep learning applications to genomics: variant intepretation

6)      deep learning applications to genomics: single cell RNAseq analysis and interpretation

7)      multiomics data integration: (MOFA hands-on in R or Python)

8)      additional hands-on sessions using R and Python workflows for the analysis and classification of omics data.

Seminars in Bioinformatics and Machine Learning

Ad hoc selected scientific papers