Analysis of real functions of real variable, in both single and multivariable cases.
Basics of linear algebra.
Basics of statistics.
Concepts of dynamical systems, both in continuous and discrete time.
At the end of the module, the student has knowledge of a wide range of problems typical of the data science and automation settings. In particular, the student is able to: formulate a business problem as a data science problem, using the most suitable technique to solve them; formulate and solve regression and classification problems; solve problems of image analysis and object recognition; apply clustering and dimensionality reduction techniques; estimate real-time non-measurable physical quantities; formulate and solve a fault diagnosis problem in an industrial environment; program a PLC to control industrial machinery.
The course will provide both theoretical lessons and laboratory experience (especially a hands-on lab with real PLCs). Matlab and python code will be given to apply the theoretical concepts.
**PART I: DATA SCIENCE**
Introduction to data science
The business perspective and the CRISP-DM process
Supervised vs. Unsupervised problems
Linear regression
Feasibility of learning
Bias-Variance tradeoff
Logistic regression
Overfitting and regularization
Validation and cross-validation
Performance metrics
Decision trees
Neural networks
Machine vision: classic approaches
Convolutional Neural Networks & deep learning
Object detection
Unsupervised learning
k-means clustering
Principal Component Analysis (PCA)
Kalman Filter
Extended Kalman Filter
Introduction to fault diagnosis
Model-based FD: parity space\observer approach
Signal-based FD: bearing inner race pitting with vibration data
Data-driven FD: Statistical Process Monitoring with T^2 e Q statistics
**PART II: AUTOMATION**
Introduction to industrial automation
Introduction to PLC
Ladder language
Structured text language
Automatic PLC code generation
Laboratory experience
- Provost, Foster, and Tom Fawcett. Data Science for Business: What you need to know about data mining and data-analytic thinking. " O'Reilly Media, Inc.", 2013.
- James, Gareth, et al. An introduction to statistical learning. Vol. 112. New York: springer, 2013.
- Verhaegen, Michel, and Vincent Verdult. Filtering and system identification: a least squares approach. Cambridge university press, 2007.
- B. M. Wilamowski, J. D. Irwin, Industrial Communications Systems, 1st Ed., Editore: CRC Press, Anno edizione: 2017, ISBN: 9781138071803
Frontal lessons with slides. The slides will be provided before the lessons.
- Matlab and Python code for the practical implementation of the concepts seen.
- Laboratory experience with real PLCs
Written exam