
Course unit
LABORATORY OF COMPUTATIONAL PHYSICS (MOD. B)
SCP8082526, A.A. 2019/20
Information concerning the students who enrolled in A.Y. 2019/20
Integrated course for this unit
ECTS: details
Type 
ScientificDisciplinary Sector 
Credits allocated 
Core courses 
FIS/01 
Experimental Physics 
6.0 
Course unit organization
Period 
Annual 
Year 
1st Year 
Teaching method 
frontal 
Type of hours 
Credits 
Teaching hours 
Hours of Individual study 
Shifts 
Laboratory 
3.0 
24 
51.0 
4 
Lecture 
3.0 
24 
51.0 
No turn 
Common characteristics of the Integrated Course unit
Prerequisites:

Even though not strictly required, the development of the class assumes the attendance of at least two physics laboratory classes during the bachelor degree 
Target skills and knowledge:

The didactic objective of this class is to teach main data analysis techniques and their application to solve concreate physics problems.
The lectures will review the main methods to extract information from complex physics datasets. The students will be able to gather, summarise and visualise the statistically relevant features of a dataset; furthermore they will learn how to qualitatively and critically compare theoretical predictions with the experimental data.
That knowledge will have to be exercised on practical lab tests, devoted to the analysis of datasets relevant to various scientific areas, i.e. biophysics, astronomy, high energy physics, etc. 
Examination methods:

To verify the proficiency of the students in the subjects covered by this course, the written reports on the lab experiences will be evaluated; such evaluation will have to be confirmed by an oral exam, during which the students will also be interviewed about what is thought during the lectures.
The oral exam will be split into two parts, each relevant to one of the two modules the class consists of. 
Assessment criteria:

The written reports on the lab experiences will have to respect the standards of a scientific publication. The data analysis will have to be tailored to the actual scientific problem being tackled and will have to demonstrate originality and the mastering of the established methodology. During the oral exam, in addition to the critical review of the written reports, the comprehension of the fundamental concepts will be tested 
Specific characteristics of the Module
Course unit contents:

1. Introduction. BiasVariance decomposition
2. Gradient descent methods
3. Linear regression: Ridge and LASSO
4. Logistic regression
5. Combining models: bagging, boosting, and random forests
6. Feedforward deep neural networks: basics
7. Deep neural networks: regularization
8. Deep neural networks: examples
9. Clustering
10. Energybased models
11. Restricted Boltzmann machines
12. Concluding examples 
Planned learning activities and teaching methods:

The aim of this course is to expose the students to modern tools for classifying data and machine learning techniques, so that they can apply those methods in lab experiences with computers. The first half of the course is reserved for this purpose of learning general principles via applications, while the second half of the course allows the students, in small groups, to develop a deeper understanding of one specific subject by carrying out a small project.
Each lesson of the first half of the course will include first a theoretical explanation of a key procedure for data analysis or of a class of algorithms, and a second phase in which the students will apply the new ideas on computers. This learning by practical experience is expected to improve the understanding of the theoretical tools and of course it is in line with the classic methodology of lab teaching. The numerical analysis includes either adopting and modifying prebuilt software, or sketching simple algorithms from scratch.
The text mainly followed in the course is an open access review on the arxiv:
“A highbias, lowvariance introduction to Machine Learning for physicists” by Pankaj Mehta et al, arXiv:1803.08823.
This review also furnishes useful python notebooks to analyze data and is connected to tools as the scikitlearn package. 
Textbooks (and optional supplementary readings) 

Pankaj Mehta, Marin Bukov, ChingHao Wang, Alexandre G.R. Day, Clint Richardson, Charles K. Fisher,, “A highbias, lowvariance introduction to Machine Learning for physicists”. : , . review avaliable open access online: https://arxiv.org/abs/1803.08823

Innovative teaching methods: Teaching and learning strategies
 Lecturing
 Laboratory
 Problem based learning
 Working in group
 Problem solving
Innovative teaching methods: Software or applications used
 Moodle (files, quizzes, workshops, ...)
 python

