Course wrap up J uly 26, 2005 CS 486/ 686 Universit y of Wat - - PowerPoint PPT Presentation

Course wrap up

J uly 26, 2005 CS 486/ 686 Universit y of Wat erloo

CS486/686 Lecture Slides (c) 2005 P. Poupart

2

Out line

• Course wrap up
• Final exam inf o
• Ot her AI courses
• AI j obs
• AI research

CS486/686 Lecture Slides (c) 2005 P. Poupart

3

Agent s and Environment s

environment percept s

act ions

? agent sensors

act uat ors Agent s include humans, robot s, sof t bot s, t hermost at s… The agent f unct ion maps percept s t o act ions f :P* A The agent pr ogram runs on t he physical archit ect ure t o produce f

CS486/686 Lecture Slides (c) 2005 P. Poupart

4

Rat ional Agent s

• Recall: A rat ional agent “does t he right t hing”
• Perf ormance measure – success crit eria

– Evaluat es a sequence of environment st at es

• A r at ional agent chooses whichever act ion

maximizes t he expect ed value of it s perf ormance measure given t he percept sequence t o dat e

– Need t o know perf ormance measure, environment , possible act ions, percept sequence

• Rat ionalit y ≠ Omniscience, Perf ect ion, Success
• Rat ionalit y explorat ion, learning, aut onomy

CS486/686 Lecture Slides (c) 2005 P. Poupart

5

Bounded Rat ionalit y

• What if t he best st rat egy given past

percept s cannot be implement ed wit h t oday’s comput ers?

• We have seen many t heories f or

rat ional agent s but what if t hose t heories are int ract able?

• Bounded rat ionalit y: f ind best

implement able st rat egy given past percept s

CS486/686 Lecture Slides (c) 2005 P. Poupart

6

Ot her AI courses

• CS498: Machine Learning – St at ist ical and

Comput at ional Foundat ions

• CS498: I mage and vision comput ing
• CS785: I nt elligent Comput er I nt eract ion
• CS886: Topics in AI : Reasoning under

Uncert aint y

• CS886: Topics in AI : Knowledge

represent at ion

CS486/686 Lecture Slides (c) 2005 P. Poupart

7

CS498: Machine Learning – St at ist ical and Comput at ional Foundat ions

• I nst ruct or: Shai Ben David
• Term: Wint er 2006
• Obj ect ives:

– The course is aimed t o f amiliarize t he st udent s wit h t he basic t heoret ical t ools and issues underlying some of t he most usef ul machine learning t echniques. The t heory of machine learning draws f rom several est ablished mat hemat ical areas including st at ist ics, geomet ry, combinat orics, and comput at ional complexit y.

CS486/686 Lecture Slides (c) 2005 P. Poupart

8

CS785 I nt elligent Comput er I nt eract ion

• I nst ruct or: Robin Cohen
• Term: Fall 2005 or Spring 2006
• Topics:

– mult iagent syst ems, – int elligent t ut oring syst ems and knowledge-based syst ems, – dat amining, – user modeling, – nat ural language generat ion and dialogue, – plan recognit ion

CS486/686 Lecture Slides (c) 2005 P. Poupart

9

CS886: Topics in AI : Reasoning under Uncert aint y

• I nst ruct or: P

ascal Poupart

• Term: Fall 2005
• Obj ect ives:

– This course will f ocus on t he principles of probabilist ic reasoning and sequent ial decision making f or a wide range of set t ings including adapt ive and mult i-agent syst ems. The modeling t echniques t hat will be covered are quit e versat ile and can be used t o t ackle a wide range of problems in many f ields including robot ics (e.g., mobile robot navigat ion, cont rol), comput er syst ems (e.g., aut onomic comput ing, query

• pt imizat ion), human-comput er int eract ion (e.g., spoken

dialog syst ems, user modeling), bioinf ormat ics (e.g., gene sequencing, design of experiment s), oper at ions research (e.g., resource allocat ion, maint enance scheduling, planning), et c. Hence, t he course should be of int erest t o a wide audience beyond art if icial int elligence.

CS486/686 Lecture Slides (c) 2005 P. Poupart

10

CS886: Topics in AI : Knowledge Represent at ion

• I nst ruct or: Chrysanne DiMarco
• Term: Spring 2006
• Topics: TBA

CS486/686 Lecture Slides (c) 2005 P. Poupart

11

AI research group

• Web: ai.uwat erloo.ca
• Prof essors:

– Shai Ben David (learning t heory) – Chrysanne DiMarco (nat ural language processing) – Pet er Van Beek (const raint programming) – Robin Cohen (mult i-agent syst ems, user modeling) – Pascal Poupart (reasoning under uncert aint y, machine learning) – Kat e Larson (game t heory, mechanism design) – Richard Mann (comput at ional vision)

CS486/686 Lecture Slides (c) 2005 P. Poupart

12

AI j obs

• Very f ew “AI companies”
• AI t ends t o be embedded in many applicat ions
• Many companies have AI R&D groups

– I nt el, Microsof t , I BM, Google, NEC, Yahoo, HP

• AI is a growing indust r y
• Has t he pot ent ial t o revolut ionize t he

comput er indust ry!