The course will allow students to deepen the methods and algorithms typical of Machine Learning (supervised, unsupervised and with reinforcement) and to use them as tools for the development of innovative technologies. In particular, aspects of the main areas of the discipline will be studied, including regression, classification and clustering. The methods and techniques of deep learning and specialized development environments will then be introduced. The course includes the development of an individual or group project that will allow students to apply the theoretical foundations learned in class to concrete problems on various domains of interest. They will be related, for example, to how to analyze large and complex datasets in various fields (e.g., Health Care, Data Science, Data Mining, Financial Analysis, Videogames, Computer Vision, etc.), create systems that adapt and improve over time (e.g., Recommender Systems), and so on. Finally, the course includes monographic seminars dedicated to various case studies.
Curriculum
scheda docente
materiale didattico
- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Programma
1. Introduction to the Course- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Testi Adottati
Lecture notes by the professor.Bibliografia Di Riferimento
S.J.Russel, P.Norvig "Artificial Intelligence: A Modern Approach", 4/Ed (2020). Pearson Education.Modalità Erogazione
Lessons and exercises. Project supervision.Modalità Valutazione
Written test, project evaluation.
scheda docente
materiale didattico
- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Programma
1. Introduction to the Course- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Testi Adottati
Lecture notes by the professor.Bibliografia Di Riferimento
S.J.Russel, P.Norvig "Artificial Intelligence: A Modern Approach", 4/Ed (2020). Pearson Education.Modalità Erogazione
Lessons and exercises. Project supervision.Modalità Valutazione
Written test, project evaluation.
scheda docente
materiale didattico
- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Programma
1. Introduction to the Course- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Testi Adottati
Lecture notes by the professor.Bibliografia Di Riferimento
S.J.Russel, P.Norvig "Artificial Intelligence: A Modern Approach", 4/Ed (2020). Pearson Education.Modalità Erogazione
Lessons and exercises. Project supervision.Modalità Valutazione
Written test, project evaluation.
scheda docente
materiale didattico
- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Programma
1. Introduction to the Course- Areas of interest in Machine Learning.
- Potenzialità dei modelli e dei metodi di ML.
2. Regression
- Introduction to Linear Regression.
- Overfitting in Regression.
- Regularization: Ridge Regression.
- Feature Selection and Lasso.
3. Classification
- Logistic Regression for classification.
- Overfitting in Classification.
- Boosting. AdaBoost algorithm.
- Naïve Bayes.
- Support Vector Machines.
4. Clustering
- Algoritmi k-means e k-means++ algorithms
- Expectation Maximization.
- Clustering gerarchico.
5. Artificial Neural Networks
- Architecture of Artificial Neural Networks.
- Backpropagation Learning algorithm.
- Applications of Artificial Neural Networks.
6. Keras and TensorFlow environments
- Keras and TensorFlow languages for developing ML applications.
- GPU-based architectures. Nvidia Tesla and Volta GPUs.
- Using TensorFlow with GPU support.
7. Dimensionality Reduction
- Data compression and visualization.
- Principal Component Analysis (PCA).
- Choosing the number of principal components.
- Applications in Recommender Systems.
8. Reinforcement Learning
- Markov Decision Process.
- Dynamic Programming.
- Reinforcement Learning algorithms.
9. Introduction to Deep Learning
- Introduction to Deep Forward Networks.
- Notes on Convolutional Neural Networks (CNN).
- Notes on Generative Adversarial Networks (GAN).
10. Case Studies and Projects
Various case studies will be presented and projects will be proposed in which to apply the notions learned on various domains of interest. In particular, the topics covered may concern, among other things, applications of ML methods and techniques in the following areas:
• Social Media Analysis (sentiment analysis, fake news detection, fake users detection, ecc.)
• Financial Machine Learning (algorithmic trading, ecc.)
• Recommender Systems (social RecSys, cultural heritage RecSys, e-commerce RecSys, ecc.)
• Data Science (prediction functions per applicazioni pratiche, ecc.)
• Computer Vision (object detection, face detection, face recognition, content-based video analysis, ecc.)
• Bioinformatics (recognition of genetic sequences, ecc.)
Testi Adottati
Lecture notes by the professor.Bibliografia Di Riferimento
S.J.Russel, P.Norvig "Artificial Intelligence: A Modern Approach", 4/Ed (2020). Pearson Education.Modalità Erogazione
Lessons and exercises. Project supervision.Modalità Valutazione
Written test, project evaluation.