Courses of Academic Year 2019/2020

Data Science and Machine Learning: basics and applications to Health Care

Lecturer: Dr. Paolo Missier

Schedule:

Le lezioni di novembre (PART I) si tengono in sala master nei seguenti orari:
- Lunedì 25 dalle 14 alle 19
- Mercoledì 27 dalle 9 alle 13
- Giovedì 28 dalle 15 alle 19

PART II date e orari da definire, settimana 13-17 gennaio

Program:

PART I Fundamentals of Machine Learning methods, overview of Data Science in the Health Care

1. Introduction to Data Science and the role of Machine Learning: Exploratory vs Predictive Data Analytics (EDA, PDA) ?
2. EDA: Overview of common techniques with notebook examples ?
3. Machine Learning: basic techniques, common pitfalls, and how to avoid them ?
4. Expressive intelligible supervised learning models: General Additive models with pairwise interactions (GA2M). Shaply index ?
5. Application. Insights from wearable activity trackers: Human Activity Recognition using a public benchmark dataset ?

PART II Health Applications and transition to Deep Learning

• The UK Biobank: opportunities for research: working with Electronic Health Records ?
• Genomics: From DNA sequencing to variant calling: a big data processing pipeline. ?
• Genome-wide Association studies (GWAS). ?
  • GWAS using the Hail platform (Spark): a complete example from the Broad InstituteFrom GWAS to machine learning for Genome-Wide Association studies ?
• Introduction to Deep Neural Networks (Deep Learning) with applications to health care

Lab activities:  

1. The course takes a very practical, hands-on approach to illustrate key concepts, with the help of python programs that show popular data analytics and ML libraries at work in detail (Pandas and Scikit-learn). These are implemented as Jupyter notebooks and made available for students to work with throughout the course. Some understanding of python programming for scientific application is desirable, but not essential. ?

1. At the start of the course, students will have the opportunity (optionally) to embark in a week-long data science experience using high frequency triaxial accelerometers (activity trackers) made available by Newcastle University. They will be able to collect their own activity data and pre-process them using open source third party SW, and then implement their owned hoc analysis algorithms to “make sense” of the activity traces. ?

Main references:

• Deep Learning book, Ian Goodfellow and Yoshua Bengio and Aaron Courville, MIT Press, 2016.  http://www.deeplearningbook.org/ ?
  • Note: we do not cover Deep Learning in this course. We only use part I of the book: "* Part I: Applied Math and Machine Learning Basics"  ?
• O’Reilly Scikit-learn book: https://www.oreilly.com/library/view/hands-on-machine-learning/9781492032632/ ?
• scikit-learn user guide, Release 0.21.3 https://scikit-learn.org/stable/_downloads/scikit-learn-docs.pdf ?

Exam:
Not yet defined

CFD: 8

Corsi Di Formazione Complementare Per Dottorandi E Assegnisti Ediz. 2019

Schedule:
18-21 November 2019

Program:
See program details .
La partecipazione all'evento sarà possibile previa iscrizione, entro e non oltre la data del 14.11.2018, tramite il modulo disponibile al link: https://docs.google.com/forms/d/e/1FAIpQLSflg423UJTyqwsH5T0dl-tFvzQPXpWwVkm9bAZWvWlfmex01A/viewform?usp=sf_link

Le presentazioni illustrate in occasione del corso in oggetto sono disponibili a questo URL

Exam:
Not yet defined

CFD: 6

Diritti & Doveri & Pubblicazioni: Incontri con gli studenti delle Scuole di Dottorato ICT e Industriale e del Territorio

Lecturer: Pola Michele e Simona Assirelli

Schedule:
lun 09/03 e merc 11/03 - ore 14.00 - 17.00

Program:
See program and schedule

Exam:
Not yet defined

CFD: 2

Modern approaches to Entity Resolution

Lecturer: Donatella Firmani (Roma Tre University)

Schedule:
June 15th 2020 - June 24th 2020

Program:
Entity Resolution (ER) is a fundamental Data Integration task. ER seeks to identify and match different manifestations of the same real-world object over different data sources, i.e., duplicates. The sheer number of ways in which real-world entities can be represented and misrepresented make ER a challenging task for automated strategies, but relatively easy for expert humans. The course aims at illustrating recent frameworks for leveraging and representing human knowledge in the context of ER -- ranging from crowdsourcing approaches to machine learning methods -- and related notions (such as, explainable AI methods and knowledge graphs).

More on the course website http://inf.uniroma3.it/~firmani/coursephd/

Exam:
individual project

CFD: 3

Academic English Workshop I

Lecturer: Silvia Cavalieri

Schedule:

• 22/6 9-12
• 24/6 9-12
• 1/7 9-12
• 3/7 9-12

Link zoom: https://univr.zoom.us/j/3553220163

Program:
The workshop aims at giving an overview of the linguistic conventions adhered to by the English-speaking academic community, focussing on aspects such as the structure of research articles, the writing of abstracts and the preparation of conference presentations. More specifically, the workshop will try to raise the participants' awareness of the rhetorical and discourse patterns characteristic of academic English, introducing them to the skills required to produce texts which are accurate both from a grammatical and a stylistic point of view. Ample opportunities will be given for practice, both in the analysis and the production of texts. In particular, special attention will be given to argumentative writing techniques and to special genres such as the abstract and conference presentations.

Exam:
Written assessment: abstract writing; oral assessment: conference presentation.

CFD: 3

Academic English Workshop II

Lecturer: Silvia Cavalieri

Schedule:

• 15/7 9.00-12.00
• 16/7 9.00-12.00
• 17/7 9.00-12.00
• 21/7 14.30-16.30

Program:
The main objective of the workshop is to provide PhD students with the knowledge of the rhetorical and linguistic models that characterize the English academic language. The seminar intends to introduce language and stylistic tools t to write accurate texts. In particular, the AEWII will focus on some specific written genres of the academia, i.e. doctoral theses and posters and Phd students will have extensive opportunities to practice both with text analysis and with their production. Special attention will be given to argumentative techniques of posters ‘ oral presentations.

Exam:
written assessment: writing a paper introduction; oral assessment: poster presentation.

CFD: 3

Digital Humanities and Digital Communication: Multimodality and Transmediality

Schedule:
September – December 2020

Program:
see here

Exam:
Not yet defined

CFD: 0