Return-Path: Delivered-To: cadet@irif.fr Received: from mailhub.math.univ-paris-diderot.fr ([81.194.30.253]) by mailhost.irif.fr (Dovecot) with LMTP id KaPkDunEKWP9UgAAP9ZUWg for ; Tue, 20 Sep 2022 15:49:29 +0200 Received: from mailhub.math.univ-paris-diderot.fr (localhost [127.0.0.1]) by mailhub.math.univ-paris-diderot.fr (Postfix) with ESMTP id 3B958F33EA for ; Tue, 20 Sep 2022 15:49:29 +0200 (CEST) X-Virus-Scanned: amavisd-new at math.univ-paris-diderot.fr X-Spam-Flag: NO X-Spam-Score: -3.048 X-Spam-Level: X-Spam-Status: No, score=-3.048 tagged_above=-10000 required=5 tests=[BAYES_00=-1.9, DCC_REPUT_00_12=-0.4, HEADER_FROM_DIFFERENT_DOMAINS=0.25, HTML_MESSAGE=0.001, MAILING_LIST_MULTI=-1, T_KAM_HTML_FONT_INVALID=0.01, T_SCC_BODY_TEXT_LINE=-0.01, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no Received: from mailhub.math.univ-paris-diderot.fr ([127.0.0.1]) by mailhub.math.univ-paris-diderot.fr (mailhub.math.univ-paris-diderot.fr [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id AQSst-ADeBwV for ; Tue, 20 Sep 2022 15:49:27 +0200 (CEST) Received: from korolev.univ-paris7.fr (korolev.univ-paris7.fr [194.254.61.138]) by mailhub.math.univ-paris-diderot.fr (Postfix) with ESMTPS id 35286F33D2 for ; Tue, 20 Sep 2022 15:49:27 +0200 (CEST) Received: from smtp1-out.u-paris.fr (smtp1-out.u-paris.fr [195.220.128.196]) by korolev.univ-paris7.fr (8.14.4/8.14.4/relay1/82085) with ESMTP id 28KDnRvk026119 for ; Tue, 20 Sep 2022 15:49:27 +0200 Received: from smtp.u75.fr (smtp.u75.fr [100.65.0.1]) by mx1.u-paris.fr (Postfix) with ESMTP id D63C2CC0912; Tue, 20 Sep 2022 15:49:26 +0200 (CEST) Received: from sympa-prod1.si.u75.fr (sympa-prod1.si.u75.fr [100.65.12.29]) by smtp.u75.fr (Postfix) with ESMTP id 178F9E0608; Tue, 20 Sep 2022 15:49:26 +0200 (CEST) Received: by sympa-prod1.si.u75.fr (Postfix, from userid 110) id 0571E25B1B; Tue, 20 Sep 2022 15:49:25 +0200 (CEST) Received: from smtp1-out.u-paris.fr (smtp1-out.u-paris.fr [195.220.128.196]) by sympa-prod1.si.u75.fr (Postfix) with ESMTP id 1C755225DE for ; Tue, 20 Sep 2022 15:48:55 +0200 (CEST) Received: from smtp.u75.fr (smtp.u75.fr [100.65.0.1]) by mx1.u-paris.fr (Postfix) with ESMTP id F3D7BCC0912 for ; Tue, 20 Sep 2022 15:48:54 +0200 (CEST) Received: from u-paris.fr (www-01.u75.fr [100.65.0.21]) by smtp.u75.fr (Postfix) with ESMTP id D838CE05C0 for ; Tue, 20 Sep 2022 15:48:54 +0200 (CEST) Message-ID: <77eec12e4284204aacbd446f2ff13ffe@u-paris.fr> Date: Tue, 20 Sep 2022 13:48:54 +0000 From: DIIP Reply-To: DIIP To: Newsletter Diip MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="_=_swift_v4_1663681734_a57c0dfb1e35c443914a8e1f9497cd54_=_" List-Unsubscribe: X-Validation-by: eva.lancelin@parisdescartes.fr Subject: [newsletter.diip] diiP Seminar announcement: October 19th X-Loop: newsletter.diip@listes.u-paris.fr X-Sequence: 25 Errors-To: newsletter.diip-owner@listes.u-paris.fr Precedence: list Precedence: bulk Sender: newsletter.diip-request@listes.u-paris.fr X-no-archive: yes List-Id: List-Help: List-Subscribe: List-Unsubscribe: List-Post: List-Owner: List-Archive: Archived-At: X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.2.7 (korolev.univ-paris7.fr [194.254.61.138]); Tue, 20 Sep 2022 15:49:27 +0200 (CEST) X-Miltered: at korolev with ID 6329C4E7.001 by Joe's j-chkmail (http : // j-chkmail dot ensmp dot fr)! X-j-chkmail-Enveloppe: 6329C4E7.001 from smtp1-out.u-paris.fr/smtp1-out.u-paris.fr/null/smtp1-out.u-paris.fr/ X-j-chkmail-Score: MSGID : 6329C4E7.001 on korolev.univ-paris7.fr : j-chkmail score : . : R=. U=. O=. B=0.000 -> S=0.000 X-j-chkmail-Status: Ham --_=_swift_v4_1663681734_a57c0dfb1e35c443914a8e1f9497cd54_=_ Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable [Ouvrir cet email dans votre navigateur.](https://u-paris.fr/diip?mailpoet_= router&endpoint=3Dtrack&action=3Dclick&data=3DWyI1MTIiLCJzdGxwMmJnYmU5Y3M0a= 3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCI3MTdmNzI2YzM2ZWMiLGZhbHNlXQ) [logo--= DIIP] SEMINAR AND ON-HANDS WORKSHOP: October 19th Deep Domain Ada= ptation and Generalization Who: Dr. Shen Liang When: October 19th, 2= 022, at 4 PM (Paris Time) Where: Online (Zoom) Abstract In real= -world applications, deep learning models are often faced with challenges f= rom multi-source data with heterogeneous features. For example, in biomedic= ine, electrocardiography (ECG) signals of different patients can differ dra= stically even if they suffer from the same heart condition, thus a computer= -aided diagnosis model that works well for one patient may work poorly for = another; in astrophysics, simulation is widely used for neutrino event reco= nstruction, yet the distribution of simulated data often fails to align wit= h that of real data, thus an event reconstruction model trained on simulate= d data may not be trustworthy on real data. Two effective solutions to the = problem with multi-source data are domain adaptation and domain generalizat= ion. Domain adaptation attempts to transfer a model trained on one or multi= ple data sources to a data source where some data is already available, whi= le domain generalization attempts to generalize a model training on multipl= e data sources to unknown future data. In this seminar, I will introduce so= me of the most commonly used methodologies for domain adaptation and genera= lization, and provide suggestions on when to and when not to apply these te= chniques in the face of multi-source data. Note that this seminar requires = the audience to have basic knowledge on transfer learning and multi-task le= arning, which can be found in the seminar on June 18th. Biography = Shen Liang is a research associate at the Data Intelligence Institute of= Paris (diiP) and affiliated with the Universit=C3=A9 Paris Cit=C3=A9. He h= as worked on a variety of data management and mining problems including tim= e series analysis, semi-supervised learning, knowledge-guided deep learning= and GPU-accelerated computation within various fields such as healthcare, = manufacturing, geosciences and astrophysics. He holds a PhD in software eng= ineering from Fudan University, China. Logistics Video, slides, a= nd other materials for this lecture will be posted on [our webpage](https:/= /u-paris.fr/diip?mailpoet_router&endpoint=3Dtrack&action=3Dclick&data=3DWyI= 1MTIiLCJzdGxwMmJnYmU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCJhOGI3Y2Y4MWQ1ZD= QiLGZhbHNlXQ) after the date of the lecture. [Click here to access the= Zoom](https://u-paris.fr/diip?mailpoet_router&endpoint=3Dtrack&action=3Dcl= ick&data=3DWyI1MTIiLCJzdGxwMmJnYmU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCI0= NWUzZTYxZmMzYTYiLGZhbHNlXQ) [UniversiteParis_logoblanc](https://u-pari= s.fr/diip?mailpoet_router&endpoint=3Dtrack&action=3Dclick&data=3DWyI1MTIiLC= JzdGxwMmJnYmU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCIxMWMzNjFiZTRkN2MiLGZhb= HNlXQ) [custom](https://u-paris.fr/diip?mailpoet_router&endpoint=3Dtra= ck&action=3Dclick&data=3DWyI1MTIiLCJzdGxwMmJnYmU5Y3M0a3dzZ3M0a2Nvc29rODRvdz= RncyIsIjkiLCI2MGU5NWU5MmZlOTYiLGZhbHNlXQ) [custom](https://u-paris.fr/diip?= mailpoet_router&endpoint=3Dtrack&action=3Dclick&data=3DWyI1MTIiLCJzdGxwMmJn= YmU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCJhZmMwYmYxZDU5NDQiLGZhbHNlXQ) [cu= stom](https://u-paris.fr/diip?mailpoet_router&endpoint=3Dtrack&action=3Dcli= ck&data=3DWyI1MTIiLCJzdGxwMmJnYmU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCI5M= WQ3ODg0MGNmZjAiLGZhbHNlXQ) [custom](https://u-paris.fr/diip?mailpoet_router= &endpoint=3Dtrack&action=3Dclick&data=3DWyI1MTIiLCJzdGxwMmJnYmU5Y3M0a3dzZ3M= 0a2Nvc29rODRvdzRncyIsIjkiLCIxNDZlZjI1MDFmODEiLGZhbHNlXQ) [Se d=C3= =A9sabonner](https://u-paris.fr/diip?mailpoet_router&endpoint=3Dtrack&actio= n=3Dclick&data=3DWyI1MTIiLCJzdGxwMmJnYmU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIj= kiLCIyYmIzYThlYWVmNzEiLGZhbHNlXQ) [MailPoet](https://u-paris.fr/diip?m= ailpoet_router&endpoint=3Dtrack&action=3Dclick&data=3DWyI1MTIiLCJzdGxwMmJnY= mU5Y3M0a3dzZ3M0a2Nvc29rODRvdzRncyIsIjkiLCI3YmNmMWYwNWU0ZDUiLGZhbHNlXQ) --_=_swift_v4_1663681734_a57c0dfb1e35c443914a8e1f9497cd54_=_ Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable = = = diiP Seminar announcement: October 19th = =20 = =20 = = =
= =
<= /table>
Ouvrir cet email dans votre navigateur. =
=
= =
3D"logo-=
=
= =
= =
=

S= EMINAR AND ON-HANDS WORKSHOP: October 19th
=
=

Deep Domain Adaptation and Generalization

=
<= table style=3D"border-collapse:collapse;border-spacing:0;mso-table-lspace:0= ;mso-table-rspace:0" width=3D"100%" cellpadding=3D"0">
Who: Dr. Shen Lian= g
When: October 19th,= 2022, at 4 PM (Paris Time)
Where: Online (Zoo= m)=20

Abstract

<= /td>
In real-world applications, deep learn= ing models are often faced with challenges from multi-source data with hete= rogeneous features. For example, in biomedicine, electrocardiography (ECG) = signals of different patients can differ drastically even if they suffer fr= om the same heart condition, thus a computer-aided diagnosis model that wor= ks well for one patient may work poorly for another; in astrophysics, simul= ation is widely used for neutrino event reconstruction, yet the distributio= n of simulated data often fails to align with that of real data, thus an ev= ent reconstruction model trained on simulated data may not be trustworthy o= n real data. Two effective solutions to the problem with multi-source data = are domain adaptation and domain generalization. Domain adaptation attempts= to transfer a model trained on one or multiple data sources to a data sour= ce where some data is already available, while domain generalization attemp= ts to generalize a model training on multiple data sources to unknown futur= e data. In this seminar, I will introduce some of the most commonly used me= thodologies for domain adaptation and generalization, and provide suggestio= ns on when to and when not to apply these techniques in the face of multi-s= ource data. Note that this seminar requires the audience to have basic know= ledge on transfer learning and multi-task learning, which can be found in t= he seminar on June 18th.
= <= h2 style=3D"margin:0 0 6px;color:#222222;font-family:'open sans','helvetica= neue',helvetica,arial,sans-serif;font-size:20px;line-height:24px;margin-bo= ttom:0;text-align:left;padding:0;font-style:normal;font-weight:normal">Biog= raphy =
Shen Liang is = a research associate at the Data Intelligence Institute of Paris (diiP) and= affiliated with the Universit=C3=A9 Paris Cit=C3=A9. He has worked on a va= riety of data management and mining problems including time series analysis= , semi-supervised learning, knowledge-guided deep learning and GPU-accelera= ted computation within various fields such as healthcare, manufacturing, ge= osciences and astrophysics. He holds a PhD in software engineering from Fud= an University, China.
<= /td>

Logisti= cs

= Video, slides, and = other materials for this lecture will be posted on our webpage a= fter the date of the lecture. = <= /table> = = =
=
3D=
= =
3D"custom"=C2=A03D"custom"=C2=A03D"custom"=C2=A0=3D"custom"=C2=A0
Se d=C3=A9sabonner =
= = =
3D"MailPoet"
= = <= /table> 3D"" --_=_swift_v4_1663681734_a57c0dfb1e35c443914a8e1f9497cd54_=_--