BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Australia/Brisbane
BEGIN:STANDARD
DTSTART:19920301T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251014T105019Z
UID:EC39210A-E6B6-43B4-BF74-004BDCFEA5E7
DTSTART;TZID=Australia/Brisbane:20251007T174500
DTEND;TZID=Australia/Brisbane:20251007T200000
DESCRIPTION:Welcome to 1TJC\,\n\nFor October&#39;s journal club Andrew Smith wi
 ll be exploring the idea of foundation models for time series forecasting.
  Foundation models have been hugely successful for Natural Language Proces
 sing (NLP)\, but do they make sense for forecasting across the different d
 omains where time series forecasting is applied? What would they learn tha
 t is applicable across domains when you have 100 billion time series data 
 points to train from?\n\nMotivating this discussion is evaluating if these
  are forecasting tools that we need to pay &quot;attention&quot; to (pun intended).\
 n\nThe paper (most) under discussion is:\n\nA decoder-only foundation mode
 l for time-series forecasting\, Abhimanyu Das\, Weihao Kong\, Rajat Sen\, 
 Yichen Zhou\, International Conference on Machine Learning 2024.\nhttps://
 arxiv.org/abs/2310.10688\n\nwith attached blog:\n\nhttps://research.google
 /blog/a-decoder-only-foundation-model-for-time-series-forecasting/\n\nPlea
 se remember that as Monday is a public holiday that journal club is on the
  first work day of the week. I hope to see you all there\, and apologies f
 or the late notice.\n\nAbout the 1st Tuesday Journal-Paper Club: We meet u
 sually on the 1st Tuesday of the month as the name suggests (inspired by t
 he ABC TV series &quot;1st Tuesday Book Club&quot;). Each month\, from amongst ourse
 lves we identify a Reader Leader. He or she picks a highly cited\, &#39;top te
 n&#39; or major-prize-winning article\, preferably in an SPS or ComSoc journal
 . Through the month\, each of us reads the article. At the next meeting\, 
 the Reader Leader leads a discussion of that article\, starting with his/h
 er own appraisal. In so doing\, we all broaden our understanding of the fi
 eld and further develop a sense of community.\n\n1st Rule of 1st Tuesday J
 ournal-Paper Club: tell everyone about 1st Tuesday Journal-Paper Club.\n\n
 Co-sponsored by: Andrew Smith\n\nBrisbane German Club\, 416 Vulture Street
 \, East Brisbane\, Brisbane\, Queensland\, Australia\, 4169
LOCATION:Brisbane German Club\, 416 Vulture Street\, East Brisbane\, Brisba
 ne\, Queensland\, Australia\, 4169
ORGANIZER:awbsmith@gmail.com
SEQUENCE:40
SUMMARY:1st Tuesday Journal-Paper Club
URL;VALUE=URI:https://events.vtools.ieee.org/m/505222
X-ALT-DESC:Description: &lt;br /&gt;&lt;div class=&quot;event-description runningText&quot;&gt;\n
 &lt;p&gt;Welcome to 1TJC\,&lt;/p&gt;\n&lt;div&gt;\n&lt;p&gt;For October&#39;s journal club Andrew Smit
 h will be exploring the idea of foundation models for time series forecast
 ing. Foundation models have been hugely successful for Natural Language Pr
 ocessing (NLP)\, but do they make sense for forecasting across the differe
 nt domains where time series forecasting is applied? What would they learn
  that is applicable across domains when you have 100 billion time series d
 ata points to train from?&lt;/p&gt;\n&lt;p&gt;Motivating this discussion is evaluating
  if these are forecasting tools that we need to pay &quot;attention&quot; to (pun in
 tended).&lt;/p&gt;\n&lt;p&gt;The paper (most) under discussion is:&lt;br&gt;&lt;br&gt;&lt;em&gt;A decode
 r-only foundation model for time-series forecasting\, &lt;/em&gt;Abhimanyu Das\,
  Weihao Kong\, Rajat Sen\, Yichen Zhou\, &amp;nbsp\;International Conference o
 n Machine Learning&lt;em&gt; &lt;/em&gt;2024. &lt;br&gt;&lt;a href=&quot;https://arxiv.org/abs/2310.
 10688&quot;&gt;https://arxiv.org/abs/2310.10688&lt;/a&gt;&lt;/p&gt;\n&lt;p&gt;with attached blog:&lt;/p
 &gt;\n&lt;p&gt;https://research.google/blog/a-decoder-only-foundation-model-for-tim
 e-series-forecasting/&lt;/p&gt;\n&lt;p&gt;&amp;nbsp\;&lt;/p&gt;\n&lt;p&gt;Please remember that as Mond
 ay is a public holiday that journal club is on the first work day of the w
 eek. I hope to see you all there\, and apologies for the late notice.&lt;/p&gt;\
 n&lt;p&gt;&amp;nbsp\;&lt;/p&gt;\n&lt;/div&gt;\n&lt;p&gt;&lt;strong&gt;About the 1st Tuesday Journal-Paper Cl
 ub:&lt;/strong&gt;&amp;nbsp\;We meet usually on the 1st Tuesday of the month as the 
 name suggests (inspired by the ABC TV series &quot;1st Tuesday Book Club&quot;). Eac
 h month\, from amongst ourselves we identify a Reader Leader. &amp;nbsp\;He or
  she&amp;nbsp\;picks a highly cited\, &#39;top ten&#39; or major-prize-winning article
 \, preferably in an SPS or ComSoc journal.&amp;nbsp\;Through the month\, each 
 of us&amp;nbsp\;reads the article. At the next meeting\, the Reader&amp;nbsp\;Lead
 er leads a discussion of that article\, starting with his/her own appraisa
 l. In so doing\, we all broaden our understanding of the field and further
  develop a sense of community.&lt;/p&gt;\n&lt;p&gt;&lt;strong&gt;1st Rule of 1st Tuesday Jou
 rnal-Paper Club:&lt;/strong&gt; tell everyone about 1st Tuesday Journal-Paper Cl
 ub.&lt;/p&gt;\n&lt;p&gt;&amp;nbsp\;&lt;/p&gt;\n&lt;/div&gt;\n&lt;p&gt;&amp;nbsp\;&lt;/p&gt;
END:VEVENT
END:VCALENDAR