HC@AIxIA: AI&Health Seminar Series (2024) - MAY 9

Dear Madam/Sir,

This is to officially announce the FOURTH seminar of the "AI & Health"
series as hosted by HC@AIxIA, i.e., the "Artificial Intelligence for
Healthcare" working group of the Italian Association for Artificial
Intelligence. *Save the date: 09 MAY, 2024.*

We hope you will attend and participate in the discussion on the relevant
topics that will be presented and by our speakers.
*Feel free to share this with those potentially interested.*
Please find some details below, and a poster attached. All directions for
participating are available at https://aixia.it/gruppi/hc/.

*== Are you interested in Joining the group? ==*
Please head to https://aixia.it/en/gruppi/hc/ fo find out how. Do not
hesitate to contact us at hc-aixia@googlegroups.com for any information or
clarification.


Thank you for your interest in the AI & Health seminar series and the
HC@AIxIA working group, and see you soon!

Sincerely,
Francesco Calimeri, Mauro Dragoni, Fabio Stella
(coordinators of the HC@AIxIA working group)



*== May 2024 seminar ==*
*Link to participate: *
https://unimib.webex.com/unimib/j.php?MTID=m9525dde973507802c873fbe7d3e091f4

*2024 MAY 09 - 4:30PM CET*
*Giovanni Parmigiani*
Dana Farber Cancer Institute, Harvard T.H. Chan School of Public Health,
MA, USA

*Title*: Validation and Replicability of Prediction Algorithms in Oncology

*Abstract*: This lecture considers replicability of the performance of
predictors across studies. We suggest a general approach to investigating
this issue, based on ensembles of prediction models trained on different
studies. We quantify how the common practice of training on a single study
accounts in part for the observed challenges in replicability of prediction
performance. We also investigate whether ensembles of predictors trained on
multiple studies can be combined, using unique criteria, to design robust
ensemble learners trained upfront to incorporate replicability into
different contexts and populations. In a linear regression setting, we show
analytically and confirm via simulation that merging yields lower
prediction error than cross-study learning when the predictor-outcome
relationships are relatively homogeneous across studies. However, as
heterogeneity increases, there exists a transition point beyond which
cross-study learning outperforms merging. We provide analytic expressions
for the transition point in various scenarios and study asymptotic
properties.

*Short Bio*: Giovanni Parmigiani's research investigates statistical
principles and tools, often with a focus on understanding cancer data. He
has a long-term interest in helping families who are susceptible to
inherited cancer understand their risk and make informed decisions. He uses
Bayesian modeling and machine learning concepts to predict who is at risk
of carrying genetic variants, and to integrate literature-based and other
information about the effects of mutations. Through his three-decade long
experience implementing machine learning approaches in clinical activities,
he identified replicability across populations and heath systems as a key
roadblock to rational use of machine learning in health. He is addressing
the challenges of cross-study replication of predictions, by designing a
variety of prediction approaches that learn replicability via training on
multiple studies at once. Throughout his research activities, his broad
goals are to find innovative ways to use data science and data technologies
to fuel cancer prevention and early detection and, methodologically, to
increase the rigor end efficiency with which we leverage the vast and
complex information generated in today’s cancer research. He strives to
foster the use of data sciences as a common thread to facilitate
interactions between fields and academic cultures, and has a passion for
mentoring and training young(er) scientists in interdisciplinary settings.
He is the Associate Director for Population Sciences of the
multi-institutional Dana-Farber / Harvard Cancer Center (DF/HCC), and is
the director of the postdoctoral training grant in Quantitative Sciences
for Cancer Research at the Harvard T.H. Chan School of Public Health. His
home is in the Department of Data Science at Dana-Farber Cancer Institute,
of which he has been the Chairman from 2009 to 2018. He has also been the
faculty Leader of DF/HCC's Biostatistics and Computational Biology Program
(now Cancer Data Sciences Program) from 2009 to 2015.


[image: HC@AIxIA - Seminars AI & Health 2024 - Locandina 05.png]