|
Teacher
|
ZINGONI ANDREA
(syllabus)
• Introduction to the course (2h): – Lessons + Seminars
o Artificial intelligence definition/s and algorithm types
o Artificial intelligence history, from the first inventions to the state-of-the-art
o Uses, benefits, criticalities and issues of artificial intelligence
• Fundamentals of statistics and probability (6h) – Lessons
o Basic concepts of statistics and combinatorics
o Basic concepts of probability
o Marginal, joint and conditioned probability and events independence
o Bayes Theorem and Total Probability Theorem
o Composed and iterated experiments
o Introduction to random variables
o Continuous and discrete variables
o Distribution and probability density/mass functions
o Notable random variables
o Trasformations of random variables
o Characteristic parameters of random variables
o Systems of random variables
o Characteristic parameters of random variables systems
• Machine Learning (16h): – Lessons + Matlab and Python laboratory + Seminars
o Introduction to AI utilization
o Data inference, training, validation and testing
o Issues in using AI: bias and overfitting.
o PCA
o K-Means and Hierarchical Clustering
o Naïve Bayes classifier and anti-spam filters
o Lineare and logistic regression
o k-NN classifier
o SVM
o Decision/regression tree and Random Forest
o Boosting methods: ADABoost, Gradient Boosting Machines (XGBoost)
• Fundamentals of neural networks and deep learning (14h): Lessons + Matlab and Python laboratory + Seminars
o Introduction to alle neural networks
o Artificial neuron and perceptron
o Neural network parameters and hyperparameters
o Activation function (sigmoid, ReLu, softmax)
o Feedforward networks and training: loss function, optimization algorithm, backpropagation algorithm
o Overfitting and regularization
o Recurrent neural networks(RNN)
o Convolutional neural networks (CNN)
o Basics of the state-of-the-art deep learning techniques and of reinforcement learning.
(reference books)
- "Teoria della probabilità e variabili aleatorie, con applicazioni all’ingegneria e alle scienze", di A. Bononi e G. Ferrari, ed. Esculapio, 2008.
- "Algoritmi per I’intelligenza artificiale”, di R. Marmo, ed. Hoepli, 2020.
- "Artificial intelligence: a modern approach" 4th ed., di S. Russel, P. Norvig, ed. Global Edition.
- “Hands-on machine learning with Scikit-learn, Keras & TensorFlow” 2nd ed., di A. Géron, O’Reilly ed.
- “Neural Networks and Deep Learning: a Textbook”, di C.C. Aggarwal, ed. Springer.
|
