77 lines
3.2 KiB
Markdown
77 lines
3.2 KiB
Markdown
type:: [[meeting]]
|
||
external-links::
|
||
tags:: [[PROJECTS/PRIN-EMELIOT]]
|
||
people::
|
||
date:: [[25-01-2023]] - 14:36
|
||
isconsortium:: true
|
||
location:: Benevento
|
||
|
||
- # Agenda
|
||
- 13:00-14:00 - Pranzo
|
||
- **14:00-14:15 - Welcome, Intro (Max + Leo)**
|
||
- **14:15-16:00 - S1: Fairness**
|
||
- Fabio (Unisa)
|
||
- Giordano (Univaq)
|
||
- Open discussion: dataset/benchmark per studi su fairness, applicazioni su IoT/chatbots
|
||
- 16.00-16:30 Coffee break
|
||
- **16:30-18:00 - S2: Infrastructure and V&V**
|
||
- Felicien - IoT (Univaq)
|
||
- Fault tree analysis (Unimib, collaboration with Univaq)
|
||
- Dario - MLOps (Unisannio)
|
||
- Open discussion: fault models for ML-Intensive IoT systems V&V di ML-intensive IoT systems
|
||
- 19:30- Social dinner
|
||
- Ristorante Roseto all’Arco - Via Arco Traiano 65, Benevento
|
||
-
|
||
- ## Fabio (Unisa)
|
||
collapsed:: true
|
||
- 
|
||
- 
|
||
- Fairness is context dependent
|
||
id:: 64459a38-5ae9-4b14-9942-dadf66e8416a
|
||
- {:height 99, :width 519}
|
||
- Fairness is difficult to handle
|
||
- 
|
||
- 
|
||
- 
|
||
- Fairness with an MLOps workflow
|
||
- Mapping study
|
||
- Validate the identified practices with practitioners
|
||
- 
|
||
- Identied fairness practices
|
||
- Le pratiche di fairness riguardano applicazioni di machine learning in diversi ambiti non necessariamente in ambito software development
|
||
- Esempio di practiche
|
||
- 
|
||
- Contesto software engineering
|
||
- 
|
||
-
|
||
-
|
||
- Contesto ML
|
||
-
|
||
- 
|
||
-
|
||
- ## Giordano (Univaq)
|
||
collapsed:: true
|
||
- 
|
||
- 
|
||
- 
|
||
- 
|
||
- 
|
||
- 
|
||
- 
|
||
- 
|
||
-
|
||
- ## Felicien (Univaq)
|
||
- [CHESSIoTMBSA.mp4 - Google Drive](https://drive.google.com/file/d/1M6fTsz7-8S4Khe-FAh1Xf4FR0EW6Kde-/view)
|
||
- ## Dario
|
||
- 
|
||
- 
|
||
- 
|
||
id:: 63d15d8e-aa2b-48e5-8d42-f16974d535fc
|
||
- 
|
||
- 
|
||
- Different kinds of [[PIPELINES]]
|
||
- {:height 313, :width 519}
|
||
- 
|
||
- Platform and Infrastructure layers are impacted by the fact that we have to manage [[PIPELINES]]
|
||
-
|
||
- |