Probabilistic reasoning in inference, forecasting and decision making (1/5)

by Prof. Giulio D'Agostini (University of Roma "La Sapienza" and INFN)

Monday 7 Jun 2021, 15:00 → 17:00 Europe/Rome

Rectorate/Building-Auditorium (GSSI)

Abstract

We are often in a state of uncertainty on several issues of our interest, that might be e.g. the 'true' value to attribute to a given quantity, on the basis of the available data, or what we shall result in further observations. But, as it is well understood, we cannot only stick to certainties, since "as far as the laws of mathematics refer to reality, they are not certain, and as far as they are certain, they do not refer to reality" (A. Einstein). Instead, we have to honestly and modestly recognize that "it is scientific only to say what is more likely and what is less likely" (R. Feynman), and therefore limit ourselves to probabilistic statements concerning the possible values of a quantity or, more in general, the hypotheses of our interest, also concerning future events. This requires to abandon the idea that probability can _only_ be attached to the outcomes of repeated observations and to recover, instead, the original thoughts on the matter by Laplace and Gauss (and Bayes), which can be finally fully exploited thanks to the unprecedented computation power presently available, accompanied by progresses in mathematical methods.

General Info

This is an interdisciplinary course organized within the GSSI Astroparticle Physics PhD program (AP-INT-3) but designed and intended to be effectively attended by students (and researchers) from all the Areas, facing interdisciplinary problems and applications.

The course has a duration of 10 hours following the calendar below:

- Monday, June 7, 15:00-17:00

- Tuesday, June 8, 9:30-11:30

- Wednesday, June 9, 9:30-11:30

- Monday, June 14, 15:00-17:00

- Wednesday, June 16, 8:45-10:45

Due to Covid-19 restrictions, the access to the Auditorium will be guaranteed to the first 60 registrants, only.

Participants must wear a mask and use the allowed seats only, as to guarantee the proper distance.

 

The audience exceeding the above limits should connect by using the Zoom link below:

https://us02web.zoom.us/j/88035952084?pwd=ZDg0MHM5U0ZpRjVXTURTRyt1ZXdJUT09
Passcode: 673152

 

Registration is needed only for those planning to attend the course in the Auditorium.

aq21_01.pdf

aq21_02.pdf

aq21_03.pdf

aq21_04.pdf

aq21_05.pdf