Coalgebra Love and Beauty in Science Ana Sokolova Logic Mentoring - - PowerPoint PPT Presentation

coalgebra love and beauty in science
SMART_READER_LITE
LIVE PREVIEW

Coalgebra Love and Beauty in Science Ana Sokolova Logic Mentoring - - PowerPoint PPT Presentation

Coalgebra Love and Beauty in Science Ana Sokolova Logic Mentoring Workshop 2019 Do you know any coalgebra? Ana Sokolova LMW 2019 22-6-19 Do you know any coalgebra? Yes, you know many coalgebras ! Ana Sokolova LMW 2019 22-6-19 Some


slide-1
SLIDE 1

Coalgebra Love and Beauty in Science

Ana Sokolova

Logic Mentoring Workshop 2019

slide-2
SLIDE 2

Do you know any coalgebra?

Ana Sokolova LMW 2019 22-6-19

slide-3
SLIDE 3

Do you know any coalgebra?

Ana Sokolova

Yes, you know many coalgebras !

LMW 2019 22-6-19

slide-4
SLIDE 4

Some coalgebras

X

c

Ñ FX

Ana Sokolova LMW 2019 22-6-19

slide-5
SLIDE 5

Some coalgebras

X

c

Ñ FX

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

LMW 2019 22-6-19

slide-6
SLIDE 6

Some coalgebras

X

c

Ñ FX

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

LMW 2019 22-6-19

slide-7
SLIDE 7

Some coalgebras

X

c

Ñ FX PA X ➝ (PDX)A

x1

a

|

a

"

b

  • 2

3

|

1 3" 1 2" 1 2

| x2

a

3 x3

b

k x4

b

k

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

LMW 2019 22-6-19

slide-8
SLIDE 8

Some coalgebras

X

c

Ñ FX

Various transitions systems / automata are coalgebras

PA X ➝ (PDX)A

x1

a

|

a

"

b

  • 2

3

|

1 3" 1 2" 1 2

| x2

a

3 x3

b

k x4

b

k

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

LMW 2019 22-6-19

slide-9
SLIDE 9

Where do they live ?

X

c

Ñ FX

Ana Sokolova LMW 2019 22-6-19

slide-10
SLIDE 10

Where do they live ?

X

c

Ñ FX

Ana Sokolova

3D Organ Model 2D Tissue Model

LMW 2019 22-6-19

slide-11
SLIDE 11

Where do they live ?

X

c

Ñ FX

Ana Sokolova

3D Organ Model 2D Tissue Model

LMW 2019 22-6-19

Bartocci et al. TCS09, CAV11

slide-12
SLIDE 12

Where do they live ?

X

c

Ñ FX

Ana Sokolova

3D Organ Model 2D Tissue Model

Verification requires clear semantics

LMW 2019 22-6-19

Bartocci et al. TCS09, CAV11

slide-13
SLIDE 13

Where do they live ?

X

c

Ñ FX

Ana Sokolova

3D Organ Model 2D Tissue Model

Verification requires clear semantics and suffers from state-space explosion

LMW 2019 22-6-19

Bartocci et al. TCS09, CAV11

slide-14
SLIDE 14

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

LMW 2019 22-6-19

slide-15
SLIDE 15

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

language equivalence

LMW 2019 22-6-19

slide-16
SLIDE 16

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence

LMW 2019 22-6-19

slide-17
SLIDE 17

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

LMW 2019 22-6-19

slide-18
SLIDE 18

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

An equivalence relation R Ñ X ˆ X is a bisimulation of the NFA po, nq: X Ñ 2 ˆ pPXqA iff whenever px, yq P R, we have opxq “ opyq and for all a P A x

a

Ñ x1 ñ Dy1.y

a

Ñ y1 ^ px1, y1q P R. Bisimilarity, denoted by „, is the largest bisimulation.

LMW 2019 22-6-19

slide-19
SLIDE 19

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

LMW 2019 22-6-19

slide-20
SLIDE 20

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

R

bisimulation

LMW 2019 22-6-19

slide-21
SLIDE 21

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

R

bisimulation

LMW 2019 22-6-19

slide-22
SLIDE 22

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

R

bisimulation

a

LMW 2019 22-6-19

slide-23
SLIDE 23

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

R

bisimulation

a a

LMW 2019 22-6-19

slide-24
SLIDE 24

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

R

bisimulation

R

a a

LMW 2019 22-6-19

slide-25
SLIDE 25

Behavioural Equivalences

„ „

Ana Sokolova

NFA X ➝ 2 x (PX)A

Two states are equivalent iff the languages recognised from these two states are the same.

language equivalence bisimilarity

R

bisimulation largest bisimulation

R

a a

LMW 2019 22-6-19

slide-26
SLIDE 26

Behavioural Equivalences

„ „

Ana Sokolova

MC X ➝ DX + 1

LMW 2019 22-6-19

slide-27
SLIDE 27

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

LMW 2019 22-6-19

slide-28
SLIDE 28

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

An equivalence relation R Ñ X ˆX is a bisimulation of the MC c: X Ñ DX `1 iff whenever px, yq P R, then either cpxq “ cpyq “ ˚ or for all R-equivalence classes C we have ÿ

zPC

cpxqpzq “ ÿ

zPC

cpyqpzq. Bisimilarity, denoted by „, is the largest bisimulation.

LMW 2019 22-6-19

slide-29
SLIDE 29

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1 Why are they both called bisimilarity ?

An equivalence relation R Ñ X ˆX is a bisimulation of the MC c: X Ñ DX `1 iff whenever px, yq P R, then either cpxq “ cpyq “ ˚ or for all R-equivalence classes C we have ÿ

zPC

cpxqpzq “ ÿ

zPC

cpyqpzq. Bisimilarity, denoted by „, is the largest bisimulation.

LMW 2019 22-6-19

slide-30
SLIDE 30

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1 What do they have in common ?

An equivalence relation R Ñ X ˆX is a bisimulation of the MC c: X Ñ DX `1 iff whenever px, yq P R, then either cpxq “ cpyq “ ˚ or for all R-equivalence classes C we have ÿ

zPC

cpxqpzq “ ÿ

zPC

cpyqpzq. Bisimilarity, denoted by „, is the largest bisimulation.

LMW 2019 22-6-19

slide-31
SLIDE 31

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

LMW 2019 22-6-19

slide-32
SLIDE 32

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

R

bisimulation

LMW 2019 22-6-19

slide-33
SLIDE 33

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

R

bisimulation

LMW 2019 22-6-19

slide-34
SLIDE 34

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

μ

R

bisimulation

LMW 2019 22-6-19

slide-35
SLIDE 35

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

μ ৵

R

bisimulation

LMW 2019 22-6-19

slide-36
SLIDE 36

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

”R

μ ৵

R

bisimulation

LMW 2019 22-6-19

slide-37
SLIDE 37

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

”R

μ ৵ lifting of R to distributions

R

bisimulation

LMW 2019 22-6-19

slide-38
SLIDE 38

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

”R

μ ৵ lifting of R to distributions assign the same probability to “R-classes”

R

bisimulation

LMW 2019 22-6-19

slide-39
SLIDE 39

Behavioural Equivalences

„ „

Ana Sokolova

bisimilarity

MC X ➝ DX + 1

”R

μ ৵ lifting of R to distributions assign the same probability to “R-classes”

R

bisimulation largest bisimulation

LMW 2019 22-6-19

slide-40
SLIDE 40

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-41
SLIDE 41

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-42
SLIDE 42

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

states

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-43
SLIDE 43

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

states

  • bject in the base

category C

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-44
SLIDE 44

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

states

  • bject in the base

category C behaviour type

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-45
SLIDE 45

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

states

  • bject in the base

category C behaviour type functor on the base category C

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-46
SLIDE 46

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

states

  • bject in the base

category C behaviour type functor on the base category C form a category too

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-47
SLIDE 47

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

states

  • bject in the base

category C behaviour type functor on the base category C form a category too CoAlgCpFq

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-48
SLIDE 48

Coalgebra

Uniform framework for dynamic transition systems, based on category theory.

X

c

Ñ FX

generic notion of behavioural equivalence

«

states

  • bject in the base

category C behaviour type functor on the base category C form a category too CoAlgCpFq

Ana Sokolova

A coalgebra is generic transition system:

LMW 2019 22-6-19

slide-49
SLIDE 49

The category of F-coalgebras

Objects = coalgebras

Ana Sokolova

Arrows = coalgebra homomorphisms

CoAlgCpFq

Two states x, y P X are behaviourally equivalent, notation x « y iff there exists a coalgebra homomorphism h: X Ñ Y from c: X Ñ FX to some coalgebra d: Y Ñ FY such that hpxq “ hpyq.

LMW 2019 22-6-19

slide-50
SLIDE 50

The category of F-coalgebras

Objects = coalgebras

X

c

Ñ FX

Ana Sokolova

Arrows = coalgebra homomorphisms

CoAlgCpFq

Two states x, y P X are behaviourally equivalent, notation x « y iff there exists a coalgebra homomorphism h: X Ñ Y from c: X Ñ FX to some coalgebra d: Y Ñ FY such that hpxq “ hpyq.

LMW 2019 22-6-19

slide-51
SLIDE 51

The category of F-coalgebras

Objects = coalgebras

X

c

Ñ FX

Ana Sokolova

Arrows = coalgebra homomorphisms

CoAlgCpFq behaviour- preserving maps

Two states x, y P X are behaviourally equivalent, notation x « y iff there exists a coalgebra homomorphism h: X Ñ Y from c: X Ñ FX to some coalgebra d: Y Ñ FY such that hpxq “ hpyq.

LMW 2019 22-6-19

slide-52
SLIDE 52

The category of F-coalgebras

Objects = coalgebras

X

c

Ñ FX

Ana Sokolova

Arrows = coalgebra homomorphisms

CoAlgCpFq

h: X Ñ Y

X

h

/

cX ✏

Y

cY

✏ FX

F h

/ FY behaviour- preserving maps

Two states x, y P X are behaviourally equivalent, notation x « y iff there exists a coalgebra homomorphism h: X Ñ Y from c: X Ñ FX to some coalgebra d: Y Ñ FY such that hpxq “ hpyq.

LMW 2019 22-6-19

slide-53
SLIDE 53

The category of F-coalgebras

Objects = coalgebras

X

c

Ñ FX

Ana Sokolova

Arrows = coalgebra homomorphisms

CoAlgCpFq

h: X Ñ Y

X

h

/

cX ✏

Y

cY

✏ FX

F h

/ FY behaviour- preserving maps

Two states x, y P X are behaviourally equivalent, notation x « y iff there exists a coalgebra homomorphism h: X Ñ Y from c: X Ñ FX to some coalgebra d: Y Ñ FY such that hpxq “ hpyq.

LMW 2019 22-6-19

slide-54
SLIDE 54

Behavioural equivalence is bisimilarity

X

c

Ñ FX

Ana Sokolova LMW 2019 22-6-19

slide-55
SLIDE 55

Behavioural equivalence is bisimilarity

X

c

Ñ FX

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

LMW 2019 22-6-19

slide-56
SLIDE 56

Behavioural equivalence is bisimilarity

X

c

Ñ FX

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

„ “ «

LMW 2019 22-6-19

slide-57
SLIDE 57

Behavioural equivalence is bisimilarity

X

c

Ñ FX

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

„ “ «

LMW 2019 22-6-19

slide-58
SLIDE 58

Behavioural equivalence is bisimilarity

X

c

Ñ FX

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

„ “ « „ “ «

LMW 2019 22-6-19

slide-59
SLIDE 59

Behavioural equivalence is bisimilarity

X

c

Ñ FX PA X ➝ (PDX)A

x1

a

|

a

"

b

  • 2

3

|

1 3" 1 2" 1 2

| x2

a

3 x3

b

k x4

b

k

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

„ “ « „ “ «

LMW 2019 22-6-19

slide-60
SLIDE 60

Behavioural equivalence is bisimilarity

X

c

Ñ FX PA X ➝ (PDX)A

x1

a

|

a

"

b

  • 2

3

|

1 3" 1 2" 1 2

| x2

a

3 x3

b

k x4

b

k

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

„ “ « „ “ « „ “ «

LMW 2019 22-6-19

slide-61
SLIDE 61

Behavioural equivalence is bisimilarity

X

c

Ñ FX PA X ➝ (PDX)A

x1

a

|

a

"

b

  • 2

3

|

1 3" 1 2" 1 2

| x2

a

3 x3

b

k x4

b

k

Ana Sokolova

NFA X ➝ 2 x (PX)A

x1

a

|

a

" x2 ✏ x3

b

k ˚

MC X ➝ DX + 1

x1

1 2



1 2

  • x2

1 3 ✏ 2 3

/ x3

1

✏ x4 ✏ x5 ✏ ˚ ˚

  • n

Sets

„ “ « „ “ « „ “ «

LMW 2019 22-6-19

slide-62
SLIDE 62

Isn’t that beautiful ?

Ana Sokolova LMW 2019 22-6-19

slide-63
SLIDE 63

Isn’t that beautiful ?

Ana Sokolova LMW 2019 22-6-19

if yes, read Rutten and Jacobs!

slide-64
SLIDE 64

Isn’t that beautiful ?

Ana Sokolova LMW 2019 22-6-19

if yes, read Rutten and Jacobs! and come to my talk tomorrow at WiL

slide-65
SLIDE 65

Isn’t that beautiful ?

Ana Sokolova LMW 2019 22-6-19

if yes, read Rutten and Jacobs! and come to my talk tomorrow at WiL and to the talk of

  • ur LICS paper on

Wednesday

slide-66
SLIDE 66

Ana Sokolova LMW 2019 23-6-19

slide-67
SLIDE 67

Ana Sokolova LMW 2019 23-6-19

slide-68
SLIDE 68

Ana Sokolova

Beyond coalgebra

LMW 2019 22-6-19

slide-69
SLIDE 69

Ana Sokolova

What is the best about doing science ?

Beyond coalgebra

LMW 2019 22-6-19

slide-70
SLIDE 70

What I love about doing science

Ana Sokolova

challenging creativity freedom novelty work with people work alone beauty elegance striving for perfection discovering explaining communicating it’s rewarding relevance joy meaningful

  • pen

integrity community

LMW 2019 22-6-19

slide-71
SLIDE 71

Let’s bring some order here

Ana Sokolova

freedom beauty

challenging

striving for perfection

joy communicating

with people

alone

meaningful

relevant

  • pen

LMW 2019 22-6-19

slide-72
SLIDE 72

Let’s bring some order here

Ana Sokolova

freedom beauty

challenging

striving for perfection

joy communicating

with people

alone

meaningful

relevant

  • pen

LMW 2019 22-6-19

slide-73
SLIDE 73

Let’s bring some order here

Ana Sokolova

freedom beauty

challenging

striving for perfection

joy communicating

with people

alone

meaningful

relevant

  • pen

LMW 2019 22-6-19

slide-74
SLIDE 74

Let’s bring some order here

Ana Sokolova

freedom beauty

challenging

striving for perfection

joy communicating

with people

alone

meaningful

relevant

  • pen

LMW 2019 22-6-19

slide-75
SLIDE 75

Let’s bring some order here

Ana Sokolova

freedom beauty

challenging

striving for perfection

joy communicating

with people

alone

meaningful

relevant

  • pen

LMW 2019 22-6-19

slide-76
SLIDE 76

Let’s bring some order here

Ana Sokolova

freedom beauty

challenging

striving for perfection

joy communicating

with people

alone

meaningful

relevant

  • pen

LMW 2019 22-6-19

slide-77
SLIDE 77
  • Do what you do best and love
  • Choose relevant topics
  • Mix topics
  • Learn from masters
  • Dare to be independent
  • Leave a trace
  • There is plenty to learn: Exchange roles
  • Do not worry much
  • Have fun

How to survive the PhD time and further until tenure ?

Ana Sokolova LMW 2019 22-6-19

slide-78
SLIDE 78
  • Do what you do best and love
  • Choose relevant topics
  • Mix topics
  • Learn from masters
  • Dare to be independent
  • Leave a trace
  • There is plenty to learn: Exchange roles
  • Do not worry much
  • Have fun

How to survive the PhD time and further until tenure ?

Ana Sokolova LMW 2019 22-6-19

slide-79
SLIDE 79
  • Do what you do best and love
  • Choose relevant topics
  • Mix topics
  • Learn from masters
  • Dare to be independent
  • Leave a trace
  • There is plenty to learn: Exchange roles
  • Do not worry much
  • Have fun

How to survive the PhD time and further until tenure ?

Ana Sokolova LMW 2019 22-6-19

slide-80
SLIDE 80
  • Do what you do best and love
  • Choose relevant topics
  • Mix topics
  • Learn from masters
  • Dare to be independent
  • Leave a trace
  • There is plenty to learn: Exchange roles
  • Do not worry much
  • Have fun

How to survive the PhD time and further until tenure ?

Ana Sokolova LMW 2019 22-6-19

slide-81
SLIDE 81
  • Do what you do best and love
  • Choose relevant topics
  • Mix topics
  • Learn from masters
  • Dare to be independent
  • Leave a trace
  • There is plenty to learn: Exchange roles
  • Do not worry much
  • Have fun

How to survive the PhD time and further until tenure ?

Ana Sokolova LMW 2019 22-6-19

slide-82
SLIDE 82
  • Do what you do best and love
  • Choose relevant topics
  • Mix topics
  • Learn from masters
  • Dare to be independent
  • Leave a trace
  • There is plenty to learn: Exchange roles
  • Do not worry much
  • Have fun

How to survive the PhD time and further until tenure ?

Ana Sokolova LMW 2019 22-6-19