probability distribu ons on structured objects
play

Probability Distribu.ons on Structured Objects September - PowerPoint PPT Presentation

Probability Distribu.ons on Structured Objects September 17, 2013 Reminder HW1 is due at 11:59pm tonight There was some ambiguity in this assignment


  1. Probability ¡Distribu.ons ¡on ¡ Structured ¡Objects ¡ September ¡17, ¡2013 ¡

  2. Reminder ¡ • HW1 ¡is ¡due ¡at ¡11:59pm ¡tonight ¡ • There ¡was ¡some ¡ambiguity ¡in ¡this ¡assignment ¡ • The ¡TAs ¡gave ¡a ¡lot ¡of ¡help, ¡but ¡in ¡general, ¡ learning ¡to ¡work ¡from ¡incomplete ¡specs ¡is ¡ important ¡

  3. Probability ¡Outline ¡ • Why ¡probability? ¡ • Probability ¡review ¡ • Mul.nomials ¡vs. ¡exponen.al ¡parameteriza.on ¡ • Locally ¡vs. ¡globally ¡normalized ¡models ¡& ¡ par..on ¡func.ons ¡ • Examples ¡

  4. Why ¡Probability? ¡ • Probability ¡formalizes ¡ – The ¡concept ¡of ¡ models ¡ – The ¡concept ¡of ¡ data ¡ – The ¡concept ¡of ¡ learning ¡ – The ¡concept ¡of ¡ predic0on ¡(inference) ¡ Probability is expectation founded upon partial knowledge.

  5. Why ¡Probability? ¡ • What ¡might ¡we ¡have ¡par.al ¡knowledge ¡ about? ¡ – The ¡state ¡of ¡the ¡world ¡(test ¡data) ¡ – The ¡reliability ¡of ¡our ¡training ¡data ¡ – The ¡correctness ¡of ¡our ¡model ¡ – The ¡values ¡of ¡our ¡parameters ¡ p ( x | partial knowledge)

  6. What ¡is ¡a ¡Probability? ¡ • Limi0ng ¡(rela0ve) ¡frequency ¡of ¡events ¡ – in ¡repeated ¡(iden.cal) ¡experiments ¡ • Degree ¡of ¡belief ¡ – Subjec.ve ¡concep.on ¡ – 40% ¡chance ¡of ¡rain ¡tomorrow ¡in ¡PiXsburgh ¡ • Viewpoint ¡affects ¡ – interpreta.on ¡ – not ¡rules ¡of ¡probability ¡calculus ¡themselves ¡

  7. Discrete ¡Distribu.ons ¡ Sample ¡space ¡ x 1 x 2 Ω Discrete ¡distribu.on: ¡ ¡ ¡ ¡ ¡ ¡ ¡is ¡ finite ¡or ¡ countable , ¡ Ω but ¡no ¡bigger ¡ ¡

  8. Discrete ¡Distribu.ons ¡ f ( x ) ∈ [0 , 1] ∀ x ∈ Ω , X f ( x ) = 1 x ∈ Ω Probability ¡mass ¡func.on ¡ An ¡ event ¡is ¡a ¡subset ¡(maybe ¡one ¡element) ¡ of ¡the ¡sample ¡space, ¡ ¡ E ⊆ Ω X P ( E ) = f ( x ) x ∈ E

  9. Random ¡Variables ¡ A ¡ random ¡variable ¡ is ¡a ¡func.on ¡from ¡a ¡random ¡event ¡ from ¡a ¡set ¡of ¡possible ¡outcomes ¡( ¡ ¡ ¡) ¡and ¡a ¡probability ¡ Ω distribu.on ¡( ¡ ¡), ¡a ¡func.on ¡from ¡outcomes ¡to ¡ ρ probabili.es. ¡ Ω = { 1 , 2 , 3 , 4 , 5 , 6 } X ( ω ) = ω ( 1 if x = 1 , 2 , 3 , 4 , 5 , 6 6 ρ X ( x ) = 0 otherwise

  10. Random ¡Variables ¡ A ¡ random ¡variable ¡ is ¡a ¡func.on ¡from ¡a ¡random ¡event ¡ from ¡a ¡set ¡of ¡possible ¡outcomes ¡( ¡ ¡ ¡) ¡and ¡a ¡probability ¡ Ω distribu.on ¡( ¡ ¡), ¡a ¡func.on ¡from ¡outcomes ¡to ¡ ρ probabili.es. ¡ Ω = { 1 , 2 , 3 , 4 , 5 , 6 } ( 0 if ω ∈ { 2 , 4 , 6 } Y ( ω ) = 1 otherwise ( 1 if y = 0 , 1 2 ρ Y ( y ) = 0 otherwise

  11. Sampling ¡Nota.on ¡ x = 4 × z + 1 . 7 Expression y ∼ Distribution( θ ) Variable Distribution Random variable Parameter

  12. Sampling ¡Nota.on ¡ x = 4 × z + 1 . 7 y ∼ Distribution( θ ) Distribution Random variable Parameter

  13. Sampling ¡Nota.on ¡ x = 4 × z + 1 . 7 y ∼ Distribution( θ ) y 0 = y × x Random variable

  14. Joint ¡Probability ¡ • Probability ¡over ¡mul.ple ¡event ¡types ¡ • Tool ¡for ¡reasoning ¡about ¡dependent ¡ (correlated) ¡events ¡ A ¡ joint ¡probability ¡distribu0on ¡ is ¡a ¡probability ¡ distribu.on ¡over ¡r.v.’s ¡with ¡the ¡following ¡form: ¡  X ( ω ) � Z = Y ( ω ) ✓ �◆ ✓ �◆ x x X = 1 ≥ 0 ∀ x ∈ X , y ∈ Y ρ Z ρ Z y y x ∈ X ,y ∈ Y

  15. Joint ¡Probability ¡ • Probability ¡over ¡mul.ple ¡event ¡types ¡ • Tool ¡for ¡reasoning ¡about ¡dependent ¡ (correlated) ¡events ¡ A ¡ joint ¡probability ¡distribu0on ¡ is ¡a ¡probability ¡ distribu.on ¡over ¡r.v.’s ¡with ¡the ¡following ¡form: ¡  X ( ω ) � Words ¡ Z = Y ( ω ) Tags ¡ ✓ �◆ ✓ �◆ x x X = 1 ≥ 0 ∀ x ∈ X , y ∈ Y ρ Z ρ Z y y x ∈ X ,y ∈ Y

  16. Joint ¡Probability ¡ • Probability ¡over ¡mul.ple ¡event ¡types ¡ • Tool ¡for ¡reasoning ¡about ¡dependent ¡ (correlated) ¡events ¡ A ¡ joint ¡probability ¡distribu0on ¡ is ¡a ¡probability ¡ distribu.on ¡over ¡r.v.’s ¡with ¡the ¡following ¡form: ¡  X ( ω ) � Words ¡ Z = Y ( ω ) Trees ¡ ✓ �◆ ✓ �◆ x x X = 1 ≥ 0 ∀ x ∈ X , y ∈ Y ρ Z ρ Z y y x ∈ X ,y ∈ Y

  17. Joint ¡Probability ¡ • Probability ¡over ¡mul.ple ¡event ¡types ¡ • Tool ¡for ¡reasoning ¡about ¡dependent ¡ (correlated) ¡events ¡ A ¡ joint ¡probability ¡distribu0on ¡ is ¡a ¡probability ¡ distribu.on ¡over ¡r.v.’s ¡with ¡the ¡following ¡form: ¡  X ( ω ) � DNA ¡sequence ¡ Z = Y ( ω ) Proteins ¡ ✓ �◆ ✓ �◆ x x X = 1 ≥ 0 ∀ x ∈ X , y ∈ Y ρ Z ρ Z y y x ∈ X ,y ∈ Y

  18. Ω = { 1 , 2 , 3 , 4 , 5 , 6 } X ( ω ) = ω

  19. Ω = { 1 , 2 , 3 , 4 , 5 , 6 } X ( ω ) = ω Ω = { (1 , 1) , (1 , 2) , (1 , 3) , (1 , 4) , (1 , 5) , (1 , 6) , (2 , 1) , (2 , 2) , (2 , 3) , (2 , 4) , (2 , 5) , (2 , 6) , (3 , 1) , (3 , 2) , (3 , 3) , (3 , 4) , (3 , 5) , (3 , 6) , (4 , 1) , (4 , 2) , (4 , 3) , (4 , 4) , (4 , 5) , (4 , 6) , (5 , 1) , (5 , 2) , (5 , 3) , (5 , 4) , (5 , 5) , (5 , 6) , (6 , 1) , (6 , 2) , (6 , 3) , (6 , 4) , (6 , 5) , (6 , 6) , } X ( ω ) = ω 1 Y ( ω ) = ω 2 ( 1 if ( x, y ) ∈ Ω 36 ρ X,Y ( x, y ) = 0 otherwise

  20. Ω = { 1 , 2 , 3 , 4 , 5 , 6 } X ( ω ) = ω Ω = { (1 , 1) , (1 , 2) , (1 , 3) , (1 , 4) , (1 , 5) , (1 , 6) , (2 , 1) , (2 , 2) , (2 , 3) , (2 , 4) , (2 , 5) , (2 , 6) , (3 , 1) , (3 , 2) , (3 , 3) , (3 , 4) , (3 , 5) , (3 , 6) , (4 , 1) , (4 , 2) , (4 , 3) , (4 , 4) , (4 , 5) , (4 , 6) , (5 , 1) , (5 , 2) , (5 , 3) , (5 , 4) , (5 , 5) , (5 , 6) , (6 , 1) , (6 , 2) , (6 , 3) , (6 , 4) , (6 , 5) , (6 , 6) , } X ( ω ) = ω 1 Y ( ω ) = ω 2 ( x + y if ( x, y ) ∈ Ω 252 ρ X,Y ( x, y ) = 0 otherwise

  21. Marginal ¡Probability ¡ p ( X = x, Y = y ) = ρ X,Y ( x, y ) X p ( X = x, Y = y 0 ) p ( X = x ) = y 0 2 Y X p ( X = x 0 , Y = y ) p ( Y = y ) = x 0 2 X Ω = { (1 , 1) , (1 , 2) , (1 , 3) , (1 , 4) , (1 , 5) , (1 , 6) , (2 , 1) , (2 , 2) , (2 , 3) , (2 , 4) , (2 , 5) , (2 , 6) , (3 , 1) , (3 , 2) , (3 , 3) , (3 , 4) , (3 , 5) , (3 , 6) , X p ( X = 4 , Y = y 0 ) p ( X = 4) = (4 , 1) , (4 , 2) , (4 , 3) , (4 , 4) , (4 , 5) , (4 , 6) , y 0 2 [1 , 6] (5 , 1) , (5 , 2) , (5 , 3) , (5 , 4) , (5 , 5) , (5 , 6) , (6 , 1) , (6 , 2) , (6 , 3) , (6 , 4) , (6 , 5) , (6 , 6) , } X p ( X = x 0 , Y = 3) p ( Y = 3) = x 0 2 [1 , 6]

  22. Marginal ¡Probability ¡ Sample ¡space ¡ (NN , cat) (JJ , fuzzy) (NN , sloth) (VB , book) (NN , book) (RB , quickly)

  23. Marginal ¡Probability ¡ Sample ¡space ¡ (NN , cat) (JJ , fuzzy) (NN , sloth) (NN , · ) (VB , book) (NN , book) (RB , quickly) p ( t = NN)

  24. Marginal ¡Probability ¡ Sample ¡space ¡ (NN , cat) (JJ , fuzzy) (NN , sloth) (VB , book) (NN , book) (RB , quickly) p ( w = book)

  25. Marginal ¡Probability ¡ Sample ¡space ¡ (NN , cat) (JJ , fuzzy) (NN , sloth) (VB , book) (NN , book) (RB , quickly)

  26. Marginal ¡Probabili.es ¡ • In ¡a ¡joint ¡model ¡of ¡word ¡and ¡tag ¡sequences ¡ p( w , t ) ¡ – The ¡probability ¡of ¡a ¡word ¡sequence ¡p( w ) ¡ – The ¡probability ¡of ¡a ¡tag ¡sequence ¡p( t ) ¡ – The ¡probability ¡of ¡a ¡word ¡sequence ¡with ¡the ¡word ¡ “cat” ¡somewhere ¡in ¡it ¡ – The ¡probability ¡of ¡a ¡tag ¡sequence ¡containing ¡three ¡ verbs ¡in ¡a ¡row ¡

  27. Condi.onal ¡Probability ¡ The ¡ condi0onal ¡probability ¡ is ¡defined ¡as ¡follows: ¡ p ( X = x | Y = y ) = p ( X = x, Y = y ) = joint probability p ( Y = y ) marginal This ¡assumes ¡ p ( Y = y ) 6 = 0 We ¡can ¡construct ¡joint ¡probability ¡distribu.ons ¡out ¡of ¡ condi.onal ¡distribu.ons: ¡ p ( x | y ) p ( y ) = p ( x, y ) = p ( y | x ) p ( x )

  28. Condi.onal ¡Probability ¡Distribu.ons ¡ The ¡ condi0onal ¡probability ¡distribu0on ¡of ¡a ¡variable ¡X ¡ given ¡a ¡variable ¡Y ¡has ¡the ¡following ¡proper.es: ¡ X p ( X = x | Y = y ) = 1 ∀ y ∈ Y, x ∈ X

  29. Condi.onal ¡Probability ¡ Sample ¡space ¡ (NN , cat) (JJ , fuzzy) (NN , sloth) (VB , book) (NN , book) (RB , quickly)

  30. Condi.onal ¡Probability ¡ Sample ¡space ¡ (NN , cat) (JJ , fuzzy) (NN , sloth) (VB , book) (NN , book) (RB , quickly) p ( · | w = book)

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend