BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal iCal API//EN
X-WR-CALNAME:Events items teaser
X-WR-TIMEZONE:America/Toronto
BEGIN:VTIMEZONE
TZID:America/Toronto
X-LIC-LOCATION:America/Toronto
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20180311T070000
END:DAYLIGHT
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20040404T070000
END:DAYLIGHT
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20050403T070000
END:DAYLIGHT
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20181104T060000
END:STANDARD
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20041031T060000
END:STANDARD
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20051030T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:69f3730d2a142
DTSTART;TZID=America/Toronto:20181214T150000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20181214T150000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/phd-seminar-p
 rogressive-memory-banks-incremental-domain
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C Waterloo ON N2L 3G1 Canada
SUMMARY:PhD Seminar: Progressive Memory Banks for Incremental Domain\nAdapt
 ation
CLASS:PUBLIC
DESCRIPTION:NABIHA ASGHAR\, PHD CANDIDATE\n_David R. Cheriton School of Com
 puter Science_\n\nWe address the problem of incremental domain adaptation 
 (IDA). We\nassume each domain comes one after another\, and that we could 
 only\naccess data in the current domain. The goal of IDA is to build a\nu
 nified model performing well on all the domains that we have\nencountered.
  We propose to augment a recurrent neural network (RNN)\nwith a directly p
 arameterized memory bank\, which is retrieved by an\nattention mechanism a
 t each step of RNN transition. The memory bank\nprovides a natural way of 
 IDA: when adapting our model to a new\ndomain\, we progressively add new s
 lots to the memory bank\, which\nincreases the number of parameters\, and 
 thus the model capacity. 
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d2b3c2
DTSTART;TZID=America/Toronto:20181213T160000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20181213T160000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/phd-seminar-b
 eyond-lif-computational-power-passive-dendritic
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2310 Waterloo ON N2L 3G1 Canada
SUMMARY:PhD Seminar: Beyond LIF: The Computational Power of Passive Dendrit
 ic\nTrees
CLASS:PUBLIC
DESCRIPTION:ANDREAS STÖCKEL\, PHD CANDIDATE\n_David R. Cheriton School of 
 Computer Science_\n\nThe artificial neurons typically employed in machine 
 learning and\ncomputational neuroscience bear little resemblance to biolog
 ical\nneurons. They are often derived from the “leaky integrate and\nfir
 e” (LIF) model\, neglect spatial extent\, and assume a linear\ncombinati
 on of input variables. It is well known that these\nsimplifications have a
  profound impact on the family of functions that\ncan be computed in a sin
 gle-layer neural network. 
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d2bdb6
DTSTART;TZID=America/Toronto:20050212T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20050212T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-se
 ntiment-classification-approach-using-stacked
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: A sentiment classification approach using stacked\nsupe
 rvised learning incorporating web mined features
CLASS:PUBLIC
DESCRIPTION:Speaker: Rejean Lau (University of Alberta)\n\nSentiment classi
 fication is a form of the text categorization problem\nwhere user sentimen
 t is categorized as either positive or negative\nsentiment. It is generall
 y accepted that sentiment analysis is a more\nchallenging classification p
 roblem then topic categorization and here\nthe bag of words approach does 
 not perform as well. Using the IMDB\nsentiment dataset from Cornell Univer
 sity\, we improve on their results\nby using a stacked classifier and web-
 mined features.
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d2cca3
DTSTART;TZID=America/Toronto:20051209T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051209T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-sp
 ecialized-multi-agents-learning-system
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: Specialized multi-agents learning system
CLASS:PUBLIC
DESCRIPTION:Speaker: Nasser Mooman\n\nWith the increasing amount of availab
 le information sources for\nlearners\, the need for systems that can effec
 tively and efficiently\nmine\, retrieve\, and process such information has
  become critical.
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d2daa5
DTSTART;TZID=America/Toronto:20051130T120000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051130T120000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-co
 nvex-hidden-markov-models
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: Convex hidden Markov models
CLASS:PUBLIC
DESCRIPTION:Speaker: Linli Xu\n\nIn this talk\, I will discuss a new unsupe
 rvised algorithm for training\nhidden Markov models that is convex and avo
 ids the use of EM.
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d2e905
DTSTART;TZID=America/Toronto:20051125T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051125T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-la
 go-svm-and-rare-target-detection
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: LAGO\, SVM and rare target detection
CLASS:PUBLIC
DESCRIPTION:Speaker: Mu Zhu (Statistics and Actuarial Sciences\, University
  of\nWaterloo)\n\nI shall describe a few projects in the area of rare targ
 et detection.
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d2f6d7
DTSTART;TZID=America/Toronto:20051118T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051118T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-de
 aling-word-sense-disambiguation-lexical
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: Dealing with word sense disambiguation in lexical chain
 ing
CLASS:PUBLIC
DESCRIPTION:Speaker: Mattt Enss\n\nA lexical chain is a sequence of words i
 n a document that are\nsemantically related (i.e.\, related in meaning). L
 exical chains\nindicate where certain topics or subjects are being discuss
 ed in a\ndocument. The chains therefore can provide context and be used to
 \ndetermine where topic changes occur.
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d304b7
DTSTART;TZID=America/Toronto:20051111T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051111T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-be
 yond-integer-domains-all-different-and-global
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: Beyond integer domains: The all different and global\nc
 ardinality constraints
CLASS:PUBLIC
DESCRIPTION:Speaker: Claude-Guy Quimper\n\nAfter giving a brief summary of 
 general principles in constraint\nprogramming\, we will present two constr
 aints: the all different\nconstraint and the global cardinality constraint
 .
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d313b1
DTSTART;TZID=America/Toronto:20051104T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051104T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-st
 ructuring-interactive-cluster-analysis
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: Structuring interactive cluster analysis
CLASS:PUBLIC
DESCRIPTION:Speaker: Wayne Oldford (Dir. of Computational Math\, UW)\n\nThe
  problem of cluster analysis\, or finding groups in data\, is\ninherently 
 ill-posed\; hence the multitude of different methods which\npurport to sol
 ve \"the'' problem. In this talk\, a variety of examples\nillustrate this 
 point and cast doubt on whether a single universally\nuseful clustering me
 thod exists.
DTSTAMP:20260430T151941Z
END:VEVENT
BEGIN:VEVENT
UID:69f3730d3214d
DTSTART;TZID=America/Toronto:20051028T113000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20051028T113000
URL:https://uwaterloo.ca/artificial-intelligence-group/events/ai-seminar-wh
 os-asking-help-bayesian-approach-intelligent
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 2306C (AI lab) Waterloo ON N2L 3G1 Canada
SUMMARY:AI seminar: Who's asking for help? A Bayesian approach to intellige
 nt\nassistance
CLASS:PUBLIC
DESCRIPTION:Speaker: Bowen Hui (University of Toronto)\n\nAutomated softwar
 e customization is drawing increasing attention as a\nmeans to help users 
 deal with the scope\, complexity\, potential\nintrusiveness\, and ever-cha
 nging nature of modern software. The\nability to automatically customize f
 unctionality\, interfaces\, and\nadvice to specific users is made more dif
 ficult by the uncertainty\nabout the needs of specific individuals and the
 ir preferences for\ninteraction.
DTSTAMP:20260430T151941Z
END:VEVENT
END:VCALENDAR