Machine Learning for Industrial Engineering

Course objectives

KNOWLEDGE AND UNDERSTANDING. Through the introduction of the fundamentals on the theoretical, technical and practical aspects in the design and implementation of machine learning systems for the solution of problems regarding the analysis of signals, measurements and, more generally, of big data, based on Computational Intelligence techniques such as Bayesian learning, neural networks, fuzzy logic, evolutionary algorithms, etc., the student will reinforce the knowledge acquired in the first cycle of studies. The applications in Industrial Engineering for the solution of supervised and unsupervised problems, in particular regarding optimization, approximation, regression, interpolation, prediction, filtering, pattern recognition and classification, in order to elaborate and apply orginal ideas, will be further investigated also in a research context. CAPABILITY TO APPLY KNOWLEDGE AND UNDERSTANDING. Capability to analyze and solve problems related to the design, implementation and testing of machine learning algorithms, with particular reference to the development in Matlab/Python/TensorFlow environment, for developing machine learning solutions applied to problems of Industrial Engineering in the management, electrical, mechanical, logistics, biomedical fields and for the training of professional and business skills able to relate in the technical-scientific field of data analytics and business intelligence, in a context therefore broader than the field of Industrial Engineering. MAKING AUTONOMOUS JUDGEMENTS. The main goal of the course is to allow the student to develop machine learning systems through an appropriate formulation of the problem, a good choice of algorithms suitable for solving the problem and performing experiments in laboratory activities in order to evaluate the efficacy of the proposed solution. During the course, the main concepts and ideas that allow the effective use of machine learning algorithms in industrial applications, rather than their purely mathematical formulation, will be mainly exposed. Therefore, the student will integrate the acquired knowledge to manage the complexity of an inductive learning mechanism where new knowledge is extracted and oriented to the solution of applicatiion problems, starting from the limited information due to the organizational contingency of the course. COMMUNICATE SKILLS. The topics covered in the course are of general interest in the scientific and industrial fields, in particular in the analysis of materials, in the design of devices and circuits, in automation and control systems, in the inversion of physical and abstract models for decision-making processes, in the management of complex networks (smart grids, energy and freight distribution, biological and social networks, etc.). Nonetheless, applications of new technologies will be introduced in the development of innovative computing systems, primarily quantum computers, in which the use of computational intelligence and machine learning algorithms for the effective and cutting-edge exploitation of the same ones is essential. Following this course, the student will be able to communicate the knowledge acquired to specialists and non-specialists in the world of work and research, where she/he will develop the subsequent scientific and/or professional activities. LEARNING SKILLS. The didactic methodology implemented in the course requires an autonomous and self-managed activity of study during the development of vertical homeworks for the didactic and/or experimental investigation of some specific topics.

Channel 1
MASSIMO PANELLA Lecturers' profile

Program - Frequency - Exams

Course program
Fundamentals of Machine Learning Methods Introduction to machine learning and data-driven modeling: data preparation; generalization; regularization and structural optimization. Overview of the main clustering and classification methods. Neural networks and Deep Learning Overview of shallow feed-forward and recurrent neural networks. Introduction to Deep Learning, specific problems and solutions (double descent, vanishing/exploding gradient, barren plateau, dropout, ensembling, weight initialization). Deep feed-forward and recurrent neural networks. Generative and diffusive systems (GAN, VAE, etc.). Introduction to Hyperdimensional Computing Fundamentals of hyperdimensional computing: data representation through high-dimensional vectors and their sparse distribution properties and robustness. Vector symbolic architectures and computational models for learning and memory. Binding and superposition methods for encoding complex information. Associative memory management and hybrid Quantum-HDC approaches. HDC-based learning strategies for classification, regression, pattern recognition and eXplainable AI problems. Practical applications of HDC in the domains of industrial and information engineering, time series analysis, anomaly detection, and resource-constrained embedded systems. Introduction to Quantum Computing Introduction to quantum computing. Elementary unitary transformations, Quantum Gate Arrays. Main quantum optimization algorithms, adiabatic approaches, Quantum Approximate Optimization Algorithm (QAOA). Variational approaches, Quantum Machine Learning and Quantum Neural Networks. Case studies Quantization, classification, prediction, approximation, interpolation, optimization and filtering of industrial, logistic, energy, economic, biomedical, mechatronic and environmental data. Management of complex networks (smart grids, energy and freight distribution, biological and social networks), smart sensor networks, materials analysis, design of devices, circuits and control systems, inversion of physical models and abstract organizational and decision-making models. Interpretability of machine learning models, implementation of algorithms in contexts with limited resources, processing of natural language and biometric signals.
Prerequisites
Fundamentals of Mathematics.
Books
M. Schuld and F. Petruccione, Supervised Learning with Quantum Computers, Springer Nature, Switzerland, 2018 Teaching material provided through the course's web site.
Teaching mode
Lectures with theoretical lessons in the classroom and exercises in the laboratory. During periods when the teaching activity is suspended (due for example to causes of force majeure), remote office hour and e-learning methods replacing lectures will be activated through telematic methods, which will be promptly communicated to the students.
Frequency
Compulsory attendance is not required.
Exam mode
Oral questions on the topics of the course.
Bibliography
M. Schuld and F. Petruccione, Supervised Learning with Quantum Computers, Springer Nature, Switzerland, 2018 C.C. Aggarwal, Neural Networks and Deep Learning, Springer Cham, Switzerland, 2023 S. Haykin, Neural Networks and Learning Machines (3rd Ed.), Pearson, NJ, USA, 2009 O. Simeone, An Introduction to Quantum Machine Learning for Engineers, arXiv preprint [2205.09510], 2022
Lesson mode
Teaching in presence. In cases of force majeure (health emergencies, etc.) the remote and/or blended methods will also be activated as per current rules.
MASSIMO PANELLA Lecturers' profile

Program - Frequency - Exams

Course program
Fundamentals of Machine Learning Methods Introduction to machine learning and data-driven modeling: data preparation; generalization; regularization and structural optimization. Overview of the main clustering and classification methods. Neural networks and Deep Learning Overview of shallow feed-forward and recurrent neural networks. Introduction to Deep Learning, specific problems and solutions (double descent, vanishing/exploding gradient, barren plateau, dropout, ensembling, weight initialization). Deep feed-forward and recurrent neural networks. Generative and diffusive systems (GAN, VAE, etc.). Introduction to Hyperdimensional Computing Fundamentals of hyperdimensional computing: data representation through high-dimensional vectors and their sparse distribution properties and robustness. Vector symbolic architectures and computational models for learning and memory. Binding and superposition methods for encoding complex information. Associative memory management and hybrid Quantum-HDC approaches. HDC-based learning strategies for classification, regression, pattern recognition and eXplainable AI problems. Practical applications of HDC in the domains of industrial and information engineering, time series analysis, anomaly detection, and resource-constrained embedded systems. Introduction to Quantum Computing Introduction to quantum computing. Elementary unitary transformations, Quantum Gate Arrays. Main quantum optimization algorithms, adiabatic approaches, Quantum Approximate Optimization Algorithm (QAOA). Variational approaches, Quantum Machine Learning and Quantum Neural Networks. Case studies Quantization, classification, prediction, approximation, interpolation, optimization and filtering of industrial, logistic, energy, economic, biomedical, mechatronic and environmental data. Management of complex networks (smart grids, energy and freight distribution, biological and social networks), smart sensor networks, materials analysis, design of devices, circuits and control systems, inversion of physical models and abstract organizational and decision-making models. Interpretability of machine learning models, implementation of algorithms in contexts with limited resources, processing of natural language and biometric signals.
Prerequisites
Fundamentals of Mathematics.
Books
M. Schuld and F. Petruccione, Supervised Learning with Quantum Computers, Springer Nature, Switzerland, 2018 Teaching material provided through the course's web site.
Teaching mode
Lectures with theoretical lessons in the classroom and exercises in the laboratory. During periods when the teaching activity is suspended (due for example to causes of force majeure), remote office hour and e-learning methods replacing lectures will be activated through telematic methods, which will be promptly communicated to the students.
Frequency
Compulsory attendance is not required.
Exam mode
Oral questions on the topics of the course.
Bibliography
M. Schuld and F. Petruccione, Supervised Learning with Quantum Computers, Springer Nature, Switzerland, 2018 C.C. Aggarwal, Neural Networks and Deep Learning, Springer Cham, Switzerland, 2023 S. Haykin, Neural Networks and Learning Machines (3rd Ed.), Pearson, NJ, USA, 2009 O. Simeone, An Introduction to Quantum Machine Learning for Engineers, arXiv preprint [2205.09510], 2022
Lesson mode
Teaching in presence. In cases of force majeure (health emergencies, etc.) the remote and/or blended methods will also be activated as per current rules.
ANTONELLO ROSATO Lecturers' profile

Program - Frequency - Exams

Course program
Introduction to Machine Learning. Basic concepts on supervised and unsupervised learning. Overview of machine learning and deep learning applications. Linear regression models. Least-squares estimation. Maximum Likelihood method. Shrinkage methods (ridge regression, LASSO, elastic net). Classification methods. Linear classifiers, logistic regression, Linear Discriminant Analysis, Quadratic Discriminant Analysis, nonparametric methods. Model assessment and selection. Bias-variance decomposition. Generalization capability and generalization error. Overfitting and underfitting. Cross-validation and K-folding. Ockham's razor and early stopping. Nonlinear regression models. Polynomial models, kernel expansion, Bayesian models, neural networks, nonparametric methods. Unsupervised learning. Clustering algorithms and cluster validity methods. Fundamentals of time series prediction. Shallow neural networks. Radial Basis Function (RBF), Fuzzy Inference System (FIS) and ANFIS neurofuzzy networks, Extreme Learning Machine (ELM) and Random Vector Functional-Link (RVFL), Echo State Network (ESN). Basic concepts on randomization. Deep neural networks. Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN), Graph Neural Network (GNN). Fundamentals of hyperdimensional computing. Fundamentals of quantum computing. Quantum gates and quantum gate arrays. Quantum algorithms for optimization and information processing (QFFT, Grover, Schor). Quantum machine learning. Quantum neural networks. Hands-on practices using Matlab and Python: linear regression, overfitting and underfitting; classification and clustering; deep learning; graph neural networks; energy time series prediction; quantum computing and quantum machine learning. Applications and case studies: prediction of renewable energy sources, intelligent energy systems, smart grids; applications to real-world data (logistic, economic, biomedical, mechatronic, environmental, aerospace, etc.); behavioral analysis and biometrics; analysis of materials and industrial processes; machine learning for the IoT/IoE, cooperative and competitive multi-agent learning, smart sensor networks; federated and distributed learning systems.
Prerequisites
Fundamentals of Mathematics.
Books
I. Goodfellow, Y. Bengio, A. Courville, Deep Learning, MIT Press. Teaching material provided through the course's web site.
Frequency
Compulsory attendance is not required.
Exam mode
Oral questions on the topics of the course.
Bibliography
T. Hastie, R. Tibshirani, J. Friedman, The Elements of Statistical Learning (2nd Ed.), Springer Series in Statistics E. Alpaydin, Introduction to Machine Learning (3rd Ed.), MIT Press [author's notes] C.M. Bishop, Pattern Recognition and Machine Learning, Springer S. Theodoridis, Machine Learning: A Bayesian and Optimization Perspective, Academic Press S.O. Haykin, Neural Networks and Learning Machines (3rd Ed.), Pearson S. Theodoridis, K. Koutroumbas, Pattern Recognition (4th Ed.), Academic Press B. Kosko, Neural Networks and Fuzzy Systems: A Dynamical Systems Approach to Machine Intelligence, Prentice-Hall
Lesson mode
Teaching in presence. In cases of force majeure (health emergencies, etc.) the remote and/or blended methods will also be activated as per current rules.
ANTONELLO ROSATO Lecturers' profile

Program - Frequency - Exams

Course program
Introduction to Machine Learning. Basic concepts on supervised and unsupervised learning. Overview of machine learning and deep learning applications. Linear regression models. Least-squares estimation. Maximum Likelihood method. Shrinkage methods (ridge regression, LASSO, elastic net). Classification methods. Linear classifiers, logistic regression, Linear Discriminant Analysis, Quadratic Discriminant Analysis, nonparametric methods. Model assessment and selection. Bias-variance decomposition. Generalization capability and generalization error. Overfitting and underfitting. Cross-validation and K-folding. Ockham's razor and early stopping. Nonlinear regression models. Polynomial models, kernel expansion, Bayesian models, neural networks, nonparametric methods. Unsupervised learning. Clustering algorithms and cluster validity methods. Fundamentals of time series prediction. Shallow neural networks. Radial Basis Function (RBF), Fuzzy Inference System (FIS) and ANFIS neurofuzzy networks, Extreme Learning Machine (ELM) and Random Vector Functional-Link (RVFL), Echo State Network (ESN). Basic concepts on randomization. Deep neural networks. Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN), Graph Neural Network (GNN). Fundamentals of hyperdimensional computing. Fundamentals of quantum computing. Quantum gates and quantum gate arrays. Quantum algorithms for optimization and information processing (QFFT, Grover, Schor). Quantum machine learning. Quantum neural networks. Hands-on practices using Matlab and Python: linear regression, overfitting and underfitting; classification and clustering; deep learning; graph neural networks; energy time series prediction; quantum computing and quantum machine learning. Applications and case studies: prediction of renewable energy sources, intelligent energy systems, smart grids; applications to real-world data (logistic, economic, biomedical, mechatronic, environmental, aerospace, etc.); behavioral analysis and biometrics; analysis of materials and industrial processes; machine learning for the IoT/IoE, cooperative and competitive multi-agent learning, smart sensor networks; federated and distributed learning systems.
Prerequisites
Fundamentals of Mathematics.
Books
I. Goodfellow, Y. Bengio, A. Courville, Deep Learning, MIT Press. Teaching material provided through the course's web site.
Frequency
Compulsory attendance is not required.
Exam mode
Oral questions on the topics of the course.
Bibliography
T. Hastie, R. Tibshirani, J. Friedman, The Elements of Statistical Learning (2nd Ed.), Springer Series in Statistics E. Alpaydin, Introduction to Machine Learning (3rd Ed.), MIT Press [author's notes] C.M. Bishop, Pattern Recognition and Machine Learning, Springer S. Theodoridis, Machine Learning: A Bayesian and Optimization Perspective, Academic Press S.O. Haykin, Neural Networks and Learning Machines (3rd Ed.), Pearson S. Theodoridis, K. Koutroumbas, Pattern Recognition (4th Ed.), Academic Press B. Kosko, Neural Networks and Fuzzy Systems: A Dynamical Systems Approach to Machine Intelligence, Prentice-Hall
Lesson mode
Teaching in presence. In cases of force majeure (health emergencies, etc.) the remote and/or blended methods will also be activated as per current rules.
  • Lesson code10589741
  • Academic year2025/2026
  • CourseManagement Engineering
  • CurriculumBusiness intelligence and analytics (percorso formativo valido anche ai fini del conseguimento del doppio titolo italo-francese) - in inglese
  • Year2nd year
  • Semester1st semester
  • SSDING-IND/31
  • CFU6