2. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2
Outline
3. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
2
Outline
4. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
3. The mechanics of understanding
2
Outline
5. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
3. The mechanics of understanding
4. Kolmogorov complexity and compression
2
Outline
6. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
3. The mechanics of understanding
4. Kolmogorov complexity and compression
5. The subjects that understand
2
Outline
7. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
3. The mechanics of understanding
4. Kolmogorov complexity and compression
5. The subjects that understand
6. Handwaving
2
Outline
8. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
3. The mechanics of understanding
4. Kolmogorov complexity and compression
5. The subjects that understand
6. Handwaving
7. Phylogenies
2
Outline
9. John S Wilkins john@wilkins.id.au
1. Understanding in music, business, and science maybe
2. Traditional accounts and recent work
3. The mechanics of understanding
4. Kolmogorov complexity and compression
5. The subjects that understand
6. Handwaving
7. Phylogenies
8. Conclusions [tentative]
2
Outline
10. John S Wilkins john@wilkins.id.au
“The accompanying diagram will aid us in understanding this rather
perplexing subject.” [Darwin, Origin, chapter 4]
3
Understanding perplexing subjects
11. John S Wilkins john@wilkins.id.au
4
Biology and the big data problem
12. John S Wilkins john@wilkins.id.au
4
Biology and the big data problem
“It’s human DNA!”
13. John S Wilkins john@wilkins.id.au
“But besides this practical concern, there is a
second basic motivation for the scientific quest,
namely, man's insatiable intellectual curiosity, his
deep concern to know the world he lives in, and
to explain, and thus to understand, the unending
flow of phenomena it presents to him.”
[Carl Hempel 1962]
From the data up
14. John S Wilkins john@wilkins.id.au
“But besides this practical concern, there is a
second basic motivation for the scientific quest,
namely, man's insatiable intellectual curiosity, his
deep concern to know the world he lives in, and
to explain, and thus to understand, the unending
flow of phenomena it presents to him.”
[Carl Hempel 1962]
“Information is not knowledge.
Knowledge is not wisdom.
Wisdom is not truth.”
[Zappa 1979]
From the data up
15. John S Wilkins john@wilkins.id.au
“But besides this practical concern, there is a
second basic motivation for the scientific quest,
namely, man's insatiable intellectual curiosity, his
deep concern to know the world he lives in, and
to explain, and thus to understand, the unending
flow of phenomena it presents to him.”
[Carl Hempel 1962]
“Information is not knowledge.
Knowledge is not wisdom.
Wisdom is not truth.”
[Zappa 1979]
“Data is not information, information is not
knowledge, knowledge is not wisdom, wisdom is
not truth.”
[Robert Royar 1994]
Hempel, Carl G. "Explanation in Science and History," in
Frontiers of Science and Philosophy, ed. R.C. Colodny, 1962,
pp. 9- 19. Pittsburgh: The University of Pittsburgh Press.
Royar, Robert. “New Horizons, Clouded Vistas.” Computers
and Composition 11, no. 2 (January 1, 1994): 93–105.
Zappa, Frank. “Packard Goose”. 1979. Joe’s Garage: Acts I, II &
III. FZ Records.
From the data up
16. John S Wilkins john@wilkins.id.au
The missing level in the
DIKW pyramid
17. John S Wilkins john@wilkins.id.au
The missing level in the
DIKW pyramid
Understanding?
18. John S Wilkins john@wilkins.id.au
The missing level in the
DIKW pyramid
Understanding?
If we approach this from
the machine learning
perspective, we might
get a better idea of
human scientific
understanding
19. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
7
Traditional accounts of understanding
20. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
7
Traditional accounts of understanding
21. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
• “Empathic insight and subjective understanding provide no warrant of objective validity, no
basis for the systematic prediction or explanation of phenomena…” [Hempel 1965, 163]
7
Traditional accounts of understanding
22. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
• “Empathic insight and subjective understanding provide no warrant of objective validity, no
basis for the systematic prediction or explanation of phenomena…” [Hempel 1965, 163]
The objectivist approach requires ultimately some pragmatist notion:
7
Traditional accounts of understanding
23. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
• “Empathic insight and subjective understanding provide no warrant of objective validity, no
basis for the systematic prediction or explanation of phenomena…” [Hempel 1965, 163]
The objectivist approach requires ultimately some pragmatist notion:
• Of prediction (Trout 2002, Wittgenstein 1953 §152)
7
Traditional accounts of understanding
24. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
• “Empathic insight and subjective understanding provide no warrant of objective validity, no
basis for the systematic prediction or explanation of phenomena…” [Hempel 1965, 163]
The objectivist approach requires ultimately some pragmatist notion:
• Of prediction (Trout 2002, Wittgenstein 1953 §152)
• Of explanation (van Fraassen 1980)
7
Traditional accounts of understanding
25. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
• “Empathic insight and subjective understanding provide no warrant of objective validity, no
basis for the systematic prediction or explanation of phenomena…” [Hempel 1965, 163]
The objectivist approach requires ultimately some pragmatist notion:
• Of prediction (Trout 2002, Wittgenstein 1953 §152)
• Of explanation (van Fraassen 1980)
The utility of causal accounts ultimately constitutes understanding.
7
Traditional accounts of understanding
26. John S Wilkins john@wilkins.id.au
Since Aristotle, understanding has been seen as based on causal explanation
(cf. de Regt 2004, Baumberger et al. 2017 for summaries)
Hempel and the logical empiricists rejected subjective elements such as feelings of confidence
or enlightenment
• “Empathic insight and subjective understanding provide no warrant of objective validity, no
basis for the systematic prediction or explanation of phenomena…” [Hempel 1965, 163]
The objectivist approach requires ultimately some pragmatist notion:
• Of prediction (Trout 2002, Wittgenstein 1953 §152)
• Of explanation (van Fraassen 1980)
The utility of causal accounts ultimately constitutes understanding.
Subjectivist or phenomenological accounts of understanding are merely psychologistic on
this approach.
7
Traditional accounts of understanding
Van Fraassen, Bas C. The ScienQfic Image. Oxford: Clarendon Press, 1980.
Hempel, Carl G. Aspects of ScienQfic ExplanaQon, and Other Essays in the Philosophy of Science. New York: The Free Press, 1965.
Regt, Henk W. de. “Discussion Note: Making Sense of Understanding.” Philosophy of Science 71, no. 1 (January 1, 2004): 98–109.
Trout, J. D. “Scienffic Explanafon and the Sense of Understanding.” Philosophy of Science 69, no. 2 (June 1, 2002): 212–33.
Baumberger, Christoph, Claus Beisbart, and Georg Brun. “What Is Understanding? An Overview of Recent Debates in Epistemology and Philosophy of Science.” In Explaining
Understanding: New PerspecQves from Epistemolgy and Philosophy of Science, edited by Stephen Grimm Christoph Baumberger and Sabine Ammon, 1–34. Routledge, 2017.
27. John S Wilkins john@wilkins.id.au
Henk de Regt and his
collaborators have
developed an account
of understanding as:
8
Recent work on understanding
28. John S Wilkins john@wilkins.id.au
Henk de Regt and his
collaborators have
developed an account
of understanding as:
• Intelligibility
8
Recent work on understanding
29. John S Wilkins john@wilkins.id.au
Henk de Regt and his
collaborators have
developed an account
of understanding as:
• Intelligibility
• Empirical adequacy
8
Recent work on understanding
30. John S Wilkins john@wilkins.id.au
Henk de Regt and his
collaborators have
developed an account
of understanding as:
• Intelligibility
• Empirical adequacy
• Consistency
8
Recent work on understanding
31. John S Wilkins john@wilkins.id.au
Henk de Regt and his
collaborators have
developed an account
of understanding as:
• Intelligibility
• Empirical adequacy
• Consistency
These are contextual
features of disciplinary
or professional
understanding, without
reference to subjects
8
Recent work on understanding
De Regt, Henk W. Understanding ScienQfic Understanding. Oxford Studies in Philosophy of Science. Oxford University Press, 2017.
De Regt, Henk W., Sabine Leonelli, and Kai Eigner. ScienQfic Understanding: Philosophical PerspecQves. Piksburgh: University of
Piksburgh Press, 2009.
32. John S Wilkins john@wilkins.id.au
Henk de Regt and his
collaborators have
developed an account
of understanding as:
• Intelligibility
• Empirical adequacy
• Consistency
These are contextual
features of disciplinary
or professional
understanding, without
reference to subjects
8
Recent work on understanding
De Regt, Henk W. Understanding ScienQfic Understanding. Oxford Studies in Philosophy of Science. Oxford University Press, 2017.
De Regt, Henk W., Sabine Leonelli, and Kai Eigner. ScienQfic Understanding: Philosophical PerspecQves. Piksburgh: University of
Piksburgh Press, 2009.
34. John S Wilkins john@wilkins.id.au
1.Intelligibility: the value that scientists attribute to the cluster of qualities
of a theory (in one or more of its representations) that facilitate the use of
the theory: A scientific theory T (in one or more of its representations) is
intelligible for scientists (in context C) if they can recognize qualitatively
characteristic consequences of T without performing exact calculations.
9
De Regt’s Criteria
35. John S Wilkins john@wilkins.id.au
1.Intelligibility: the value that scientists attribute to the cluster of qualities
of a theory (in one or more of its representations) that facilitate the use of
the theory: A scientific theory T (in one or more of its representations) is
intelligible for scientists (in context C) if they can recognize qualitatively
characteristic consequences of T without performing exact calculations.
2.Empirical adequacy
Nuff said (for now; I’ll get back to this)
9
De Regt’s Criteria
36. John S Wilkins john@wilkins.id.au
1.Intelligibility: the value that scientists attribute to the cluster of qualities
of a theory (in one or more of its representations) that facilitate the use of
the theory: A scientific theory T (in one or more of its representations) is
intelligible for scientists (in context C) if they can recognize qualitatively
characteristic consequences of T without performing exact calculations.
2.Empirical adequacy
Nuff said (for now; I’ll get back to this)
3.Consistency: A phenomenon P is understood scientifically if and only if
there is an explanation of P that is based on an intelligible theory T and
conforms to the basic epistemic values of empirical adequacy and
internal consistency.
9
De Regt’s Criteria
37. John S Wilkins john@wilkins.id.au
1.Intelligibility: the value that scientists attribute to the cluster of qualities
of a theory (in one or more of its representations) that facilitate the use of
the theory: A scientific theory T (in one or more of its representations) is
intelligible for scientists (in context C) if they can recognize qualitatively
characteristic consequences of T without performing exact calculations.
2.Empirical adequacy
Nuff said (for now; I’ll get back to this)
3.Consistency: A phenomenon P is understood scientifically if and only if
there is an explanation of P that is based on an intelligible theory T and
conforms to the basic epistemic values of empirical adequacy and
internal consistency.
Note the passive nature here. It’s as if scientists don’t do very much in
understanding 9
De Regt’s Criteria
38. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
10
Formal mechanics
39. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
10
Formal mechanics
40. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
• but also because there are objections to subjectivism in psychological
approaches (with which I concur).
10
Formal mechanics
41. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
• but also because there are objections to subjectivism in psychological
approaches (with which I concur).
How to approach this?
10
Formal mechanics
42. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
• but also because there are objections to subjectivism in psychological
approaches (with which I concur).
How to approach this?
One domain in which such matters are dealt with mechanically (i.e., mechanisms are
sought) is machine learning (ML, a subset of AI).
10
Formal mechanics
43. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
• but also because there are objections to subjectivism in psychological
approaches (with which I concur).
How to approach this?
One domain in which such matters are dealt with mechanically (i.e., mechanisms are
sought) is machine learning (ML, a subset of AI).
• The mechanisms can be opaque “black boxes” in their operation (deep learning)
10
Formal mechanics
44. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
• but also because there are objections to subjectivism in psychological
approaches (with which I concur).
How to approach this?
One domain in which such matters are dealt with mechanically (i.e., mechanisms are
sought) is machine learning (ML, a subset of AI).
• The mechanisms can be opaque “black boxes” in their operation (deep learning)
• There is no subjectivity required (but there is a “subject”; that is, a learning
system)
10
Formal mechanics
45. John S Wilkins john@wilkins.id.au
The mechanism of understanding is not generally considered in epistemology…
• because we do not have the requisite neurobiological or neuropsychological
knowledge yet (we do not understand understanding);
• but also because there are objections to subjectivism in psychological
approaches (with which I concur).
How to approach this?
One domain in which such matters are dealt with mechanically (i.e., mechanisms are
sought) is machine learning (ML, a subset of AI).
• The mechanisms can be opaque “black boxes” in their operation (deep learning)
• There is no subjectivity required (but there is a “subject”; that is, a learning
system)
I will attempt to generalise ML and algorithmic information theoretic tools to apply to
this problem of understanding within knowing systems
10
Formal mechanics
46. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
11
Kinematic versus dynamic
47. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
11
Kinematic versus dynamic
48. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
• Dynamics: the relations between forces and motion*
11
Kinematic versus dynamic
*Dunamis [δύναμις] is an ordinary Greek word for possibility or capability. Depending on context, it could be
translated "potency", "potential", "capacity", "ability", "power", "capability", "strength", "possibility",
"force" [Wikipedia]
49. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
• Dynamics: the relations between forces and motion*
In brief: between phenomenal accounts and causal explanations.
11
Kinematic versus dynamic
*Dunamis [δύναμις] is an ordinary Greek word for possibility or capability. Depending on context, it could be
translated "potency", "potential", "capacity", "ability", "power", "capability", "strength", "possibility",
"force" [Wikipedia]
50. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
• Dynamics: the relations between forces and motion*
In brief: between phenomenal accounts and causal explanations.
We can explore the problem of understanding in knowing systems via the shift from
kinematic accounts (what the behaviour of the system that understands is) to dynamic
explanations (what forces the behaviour of the understanding system)
11
Kinematic versus dynamic
*Dunamis [δύναμις] is an ordinary Greek word for possibility or capability. Depending on context, it could be
translated "potency", "potential", "capacity", "ability", "power", "capability", "strength", "possibility",
"force" [Wikipedia]
51. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
• Dynamics: the relations between forces and motion*
In brief: between phenomenal accounts and causal explanations.
We can explore the problem of understanding in knowing systems via the shift from
kinematic accounts (what the behaviour of the system that understands is) to dynamic
explanations (what forces the behaviour of the understanding system)
• Of systems that are the subject of understanding
11
Kinematic versus dynamic
*Dunamis [δύναμις] is an ordinary Greek word for possibility or capability. Depending on context, it could be
translated "potency", "potential", "capacity", "ability", "power", "capability", "strength", "possibility",
"force" [Wikipedia]
52. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
• Dynamics: the relations between forces and motion*
In brief: between phenomenal accounts and causal explanations.
We can explore the problem of understanding in knowing systems via the shift from
kinematic accounts (what the behaviour of the system that understands is) to dynamic
explanations (what forces the behaviour of the understanding system)
• Of systems that are the subject of understanding
• Of systems that understand
11
Kinematic versus dynamic
*Dunamis [δύναμις] is an ordinary Greek word for possibility or capability. Depending on context, it could be
translated "potency", "potential", "capacity", "ability", "power", "capability", "strength", "possibility",
"force" [Wikipedia]
53. John S Wilkins john@wilkins.id.au
It may help if we avail ourselves of an old distinction in physics between two kinds of
description:
• Kinematics: description of motions without consideration of forces
• Dynamics: the relations between forces and motion*
In brief: between phenomenal accounts and causal explanations.
We can explore the problem of understanding in knowing systems via the shift from
kinematic accounts (what the behaviour of the system that understands is) to dynamic
explanations (what forces the behaviour of the understanding system)
• Of systems that are the subject of understanding
• Of systems that understand
These coincide: as we move from kinematic descriptions to dynamic explanations of
knowing systems, we also move from considering knowledge to considering understanding
11
Kinematic versus dynamic
*Dunamis [δύναμις] is an ordinary Greek word for possibility or capability. Depending on context, it could be
translated "potency", "potential", "capacity", "ability", "power", "capability", "strength", "possibility",
"force" [Wikipedia]
54. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
12
Pattern recognition
55. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
12
Pattern recognition
56. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
12
Pattern recognition
57. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
• E.g., regression and other statistical analyses
12
Pattern recognition
58. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
• E.g., regression and other statistical analyses
• Machine Learning classifications (ANNs)
12
Pattern recognition
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0069191
59. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
• E.g., regression and other statistical analyses
• Machine Learning classifications (ANNs)
• Distance analyses (Hamming, Manhattan)
12
Pattern recognition
60. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
• E.g., regression and other statistical analyses
• Machine Learning classifications (ANNs)
• Distance analyses (Hamming, Manhattan)
Less memory and cognitive resources are
required to retrieve and interpret patterns
12
Pattern recognition
61. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
• E.g., regression and other statistical analyses
• Machine Learning classifications (ANNs)
• Distance analyses (Hamming, Manhattan)
Less memory and cognitive resources are
required to retrieve and interpret patterns
A pattern therefore has information the data
does not
12
Pattern recognition
62. John S Wilkins john@wilkins.id.au
Knowing systems have a limit to their working memory under realistic assumptions
Knowing systems also have to deal with noisy data and uncertainty
Once the scale of the data set exceeds the capacity of the system’s cognitive limits
[working memory, processing power] pattern recognition must be out-sourced
• E.g., regression and other statistical analyses
• Machine Learning classifications (ANNs)
• Distance analyses (Hamming, Manhattan)
Less memory and cognitive resources are
required to retrieve and interpret patterns
A pattern therefore has information the data
does not
[Here, information is a property of the state
of a knowing system]
12
Pattern recognition
63. John S Wilkins john@wilkins.id.au
As any ML system is a Turing machine, it follows that, even when we do not know
it, there is an algorithm and data set A that generates the output pattern P [where
P is the phenomenon]
13
Machine learning and understanding
64. John S Wilkins john@wilkins.id.au
As any ML system is a Turing machine, it follows that, even when we do not know
it, there is an algorithm and data set A that generates the output pattern P [where
P is the phenomenon]
The information content of P is therefore the complexity of A
13
Machine learning and understanding
65. John S Wilkins john@wilkins.id.au
As any ML system is a Turing machine, it follows that, even when we do not know
it, there is an algorithm and data set A that generates the output pattern P [where
P is the phenomenon]
The information content of P is therefore the complexity of A
• If P is lossy, then its information content is lower in complexity than the data set
13
Machine learning and understanding
66. John S Wilkins john@wilkins.id.au
As any ML system is a Turing machine, it follows that, even when we do not know
it, there is an algorithm and data set A that generates the output pattern P [where
P is the phenomenon]
The information content of P is therefore the complexity of A
• If P is lossy, then its information content is lower in complexity than the data set
• If P is lossless, then the complexity of A equals the complexity of P [defn]
13
Machine learning and understanding
67. John S Wilkins john@wilkins.id.au
As any ML system is a Turing machine, it follows that, even when we do not know
it, there is an algorithm and data set A that generates the output pattern P [where
P is the phenomenon]
The information content of P is therefore the complexity of A
• If P is lossy, then its information content is lower in complexity than the data set
• If P is lossless, then the complexity of A equals the complexity of P [defn]
So A is a compression algorithm*. By analogy, when we generate a pattern (such
as a regression curve) from a large or noisy data set, we are compressing the data
to generate information, and from that information we can fairly say we know the
domain the data set samples or represents.
13
Machine learning and understanding
* The worst compression is, of course, the data set itself plus a “print” command
68. John S Wilkins john@wilkins.id.au
As any ML system is a Turing machine, it follows that, even when we do not know
it, there is an algorithm and data set A that generates the output pattern P [where
P is the phenomenon]
The information content of P is therefore the complexity of A
• If P is lossy, then its information content is lower in complexity than the data set
• If P is lossless, then the complexity of A equals the complexity of P [defn]
So A is a compression algorithm*. By analogy, when we generate a pattern (such
as a regression curve) from a large or noisy data set, we are compressing the data
to generate information, and from that information we can fairly say we know the
domain the data set samples or represents.
The complexity of P is specified in Algorithmic Information Theory (AIT) as the
length (in bits) of the shortest program that generates P.
13
Machine learning and understanding
* The worst compression is, of course, the data set itself plus a “print” command
69. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
14
Kolmogorov compressed
70. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
Kolmogorov Complexity (K) is the length of the shortest program in a language L
that generates a string w
14
Kolmogorov compressed
71. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
Kolmogorov Complexity (K) is the length of the shortest program in a language L
that generates a string w
• The most complex string in L is a fully random string
14
Kolmogorov compressed
72. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
Kolmogorov Complexity (K) is the length of the shortest program in a language L
that generates a string w
• The most complex string in L is a fully random string
• A fully random string is incompressible
14
Kolmogorov compressed
73. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
Kolmogorov Complexity (K) is the length of the shortest program in a language L
that generates a string w
• The most complex string in L is a fully random string
• A fully random string is incompressible
• The working memory of a ML system is thus the largest random string that it
can hold in memory while applying functions to it
14
Kolmogorov compressed
74. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
Kolmogorov Complexity (K) is the length of the shortest program in a language L
that generates a string w
• The most complex string in L is a fully random string
• A fully random string is incompressible
• The working memory of a ML system is thus the largest random string that it
can hold in memory while applying functions to it
Thus K gives the complexity of a string (of digits, or other structured data)
14
Kolmogorov compressed
75. John S Wilkins john@wilkins.id.au
AIT is based on the work of Andrey Kolmogorov (1903–1987) and Ray Solomonov
(1926–2009)
Kolmogorov Complexity (K) is the length of the shortest program in a language L
that generates a string w
• The most complex string in L is a fully random string
• A fully random string is incompressible
• The working memory of a ML system is thus the largest random string that it
can hold in memory while applying functions to it
Thus K gives the complexity of a string (of digits, or other structured data)
If we conceive of ourselves as knowing systems analogous to ML systems, the
information in data is thus the K “program” in our cognitive processes
14
Kolmogorov compressed
Li, Ming, and Paul Vitányi. An IntroducQon to Kolmogorov Complexity and Its ApplicaQons. 4th ed. Texts in Computer Science. Springer Internafonal
Publishing, 2019.
Grünwald, Peter, and Paul M. B. Vitányi. “Shannon Informafon and Kolmogorov Complexity.” CoRR cs.IT/0410002 (2004).
Wallace, C. S., and D. L. Dowe. “Minimum Message Length and Kolmogorov Complexity.” The Computer Journal 42, no. 4 (January 1, 1999): 270–83.
76. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
15
Reducing the complexity of data
77. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
15
Reducing the complexity of data
78. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
Complexity of the information is inversely related to the informativeness for a limited
knowing system
15
Reducing the complexity of data
79. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
Complexity of the information is inversely related to the informativeness for a limited
knowing system
• Put another way, the tractability of inferences increases as complexity reduces
15
Reducing the complexity of data
80. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
Complexity of the information is inversely related to the informativeness for a limited
knowing system
• Put another way, the tractability of inferences increases as complexity reduces
[Inversely, tractability reduces as data complexity increases, unless it is compressible]
15
Reducing the complexity of data
81. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
Complexity of the information is inversely related to the informativeness for a limited
knowing system
• Put another way, the tractability of inferences increases as complexity reduces
[Inversely, tractability reduces as data complexity increases, unless it is compressible]
• Compression of data to information is a lossy process; the structure/pattern in the
data is simplified through regression, etc.
15
Reducing the complexity of data
82. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
Complexity of the information is inversely related to the informativeness for a limited
knowing system
• Put another way, the tractability of inferences increases as complexity reduces
[Inversely, tractability reduces as data complexity increases, unless it is compressible]
• Compression of data to information is a lossy process; the structure/pattern in the
data is simplified through regression, etc.
• The function of this information is to set the expectations of the knowing system
[but that’s another talk; contrastive explanation is why]
15
Reducing the complexity of data
83. John S Wilkins john@wilkins.id.au
Note that while using complexity metrics is based on the subject – the ML or the
knowing system – it is not subjective, as the experience felt by the knower is not relevant
for understanding the information
[Nor is this an objectivist account, as it does not depend upon a truth operator]
Complexity of the information is inversely related to the informativeness for a limited
knowing system
• Put another way, the tractability of inferences increases as complexity reduces
[Inversely, tractability reduces as data complexity increases, unless it is compressible]
• Compression of data to information is a lossy process; the structure/pattern in the
data is simplified through regression, etc.
• The function of this information is to set the expectations of the knowing system
[but that’s another talk; contrastive explanation is why]
• In short: the information derived from the data sets the prior probabilities for
future data, highlighting anomalies and points of interest
15
Reducing the complexity of data
84. John S Wilkins john@wilkins.id.au
16
Compression and generalisation
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
85. John S Wilkins john@wilkins.id.au
16
Compression and generalisation
Measurement
compresses to
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
86. John S Wilkins john@wilkins.id.au
Analysis
compresses to
16
Compression and generalisation
Measurement
compresses to
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
87. John S Wilkins john@wilkins.id.au
Analysis
compresses to
16
Compression and generalisation
Measurement
compresses to
Kinematic rules
y = f(x)
applies to
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
88. John S Wilkins john@wilkins.id.au
Analysis
compresses to
Dynamic or causal models
y = f(x)
16
Compression and generalisation
Measurement
compresses to
Kinematic rules
y = f(x)
applies to
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
89. John S Wilkins john@wilkins.id.au
Analysis
compresses to
Dynamic or causal models
y = f(x)
16
Compression and generalisation
Measurement
compresses to
Kinematic rules
y = f(x)
applies to
Data Information
KnowledgeUnderstanding
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
90. John S Wilkins john@wilkins.id.au
Analysis
compresses to
Dynamic or causal models
y = f(x)
16
Compression and generalisation
Measurement
compresses to
Kinematic rules
y = f(x)
applies to
Data Information
KnowledgeUnderstanding
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
91. John S Wilkins john@wilkins.id.au
Analysis
compresses to
Dynamic or causal models
y = f(x)
16
Compression and generalisation
Measurement
compresses to
Kinematic rules
y = f(x)
applies to
Data Information
KnowledgeUnderstanding
Understanding
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
92. John S Wilkins john@wilkins.id.au
Analysis
compresses to
Dynamic or causal models
y = f(x)
16
Compression and generalisation
Measurement
compresses to
Kinematic rules
y = f(x)
applies to
Data Information
KnowledgeUnderstanding
Understanding
What counts as wisdom is left to
you to decide
“Solomonoff held that
all inference problems
can be cast in the form
of extrapolation from an
ordered sequence of
binary symbols.”
[Li and Vintányi 2019,63]
93. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
17
Subjects understanding
94. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
17
Subjects understanding
95. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
• particular scientists
17
Subjects understanding
96. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
• particular scientists
• disciplines and sub disciplines
17
Subjects understanding
97. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
• particular scientists
• disciplines and sub disciplines
• research groups
17
Subjects understanding
98. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
• particular scientists
• disciplines and sub disciplines
• research groups
I would take this approach to apply equally well to both individual and collective
knowing systems/subjects.
17
Subjects understanding
99. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
• particular scientists
• disciplines and sub disciplines
• research groups
I would take this approach to apply equally well to both individual and collective
knowing systems/subjects.
We can say that Professor X understands some phenomenon, or that Discipline D
understands some phenomenon independently (individuals can fail to understand
what the discipline does, and vice versa)
17
Subjects understanding
100. John S Wilkins john@wilkins.id.au
So, how might this analogy work in the arena of science? Scientists seek to understand
the universe, of course, as Hempel said.
• But what does “scientists” refer to? What are the knowing systems, or subjects, in
science?
• particular scientists
• disciplines and sub disciplines
• research groups
I would take this approach to apply equally well to both individual and collective
knowing systems/subjects.
We can say that Professor X understands some phenomenon, or that Discipline D
understands some phenomenon independently (individuals can fail to understand
what the discipline does, and vice versa)
[Does this mean we have actual Turing machines in our heads? No, it’s an abstract way
to consider the problem (just as neurones are not artificial neurones or v.v.)]
17
Subjects understanding
101. John S Wilkins john@wilkins.id.au
I do not have time here to give more than one proper case study; so I will handwave
at some:
18
Insert case studies here
102. John S Wilkins john@wilkins.id.au
I do not have time here to give more than one proper case study; so I will handwave
at some:
• Gas laws (Boyles’ Law, Charles’ Law, Gay-Lussac’s Law, Avogadro’s Law, etc.)
leading to the Ideal Gas Law, to van der Waals’ equation
18
Insert case studies here
103. John S Wilkins john@wilkins.id.au
I do not have time here to give more than one proper case study; so I will handwave
at some:
• Gas laws (Boyles’ Law, Charles’ Law, Gay-Lussac’s Law, Avogadro’s Law, etc.)
leading to the Ideal Gas Law, to van der Waals’ equation
• Periodic table construction by empirical lab work, followed first by valency, then
electron shell, then quantum mechanical causal accounts
18
Insert case studies here
104. John S Wilkins john@wilkins.id.au
I do not have time here to give more than one proper case study; so I will handwave
at some:
• Gas laws (Boyles’ Law, Charles’ Law, Gay-Lussac’s Law, Avogadro’s Law, etc.)
leading to the Ideal Gas Law, to van der Waals’ equation
• Periodic table construction by empirical lab work, followed first by valency, then
electron shell, then quantum mechanical causal accounts
• Mendelian genetics (phenomenal) through to population genetics (kinematic),
through to molecular genetics (dynamic)
18
Insert case studies here
105. John S Wilkins john@wilkins.id.au
I do not have time here to give more than one proper case study; so I will handwave
at some:
• Gas laws (Boyles’ Law, Charles’ Law, Gay-Lussac’s Law, Avogadro’s Law, etc.)
leading to the Ideal Gas Law, to van der Waals’ equation
• Periodic table construction by empirical lab work, followed first by valency, then
electron shell, then quantum mechanical causal accounts
• Mendelian genetics (phenomenal) through to population genetics (kinematic),
through to molecular genetics (dynamic)
• Arguably: observational to theoretical ecology, through to ecological
thermodynamics and “ecological orbits” (Haynie 2001, Ginsburg and Colyvan 2004)
18
Insert case studies here
106. John S Wilkins john@wilkins.id.au
I do not have time here to give more than one proper case study; so I will handwave
at some:
• Gas laws (Boyles’ Law, Charles’ Law, Gay-Lussac’s Law, Avogadro’s Law, etc.)
leading to the Ideal Gas Law, to van der Waals’ equation
• Periodic table construction by empirical lab work, followed first by valency, then
electron shell, then quantum mechanical causal accounts
• Mendelian genetics (phenomenal) through to population genetics (kinematic),
through to molecular genetics (dynamic)
• Arguably: observational to theoretical ecology, through to ecological
thermodynamics and “ecological orbits” (Haynie 2001, Ginsburg and Colyvan 2004)
• Tectonic drift theory (Oreskes and le Grand 2003)
18
Insert case studies here
Ginzburg, Lev R., and Mark Colyvan. Ecological Orbits : How Planets Move and PopulaQons Grow. Oxford; New York: Oxford University Press,
2004.
Haynie, Donald T. Biological Thermodynamics. Cambridge ; New York: Cambridge University Press, 2001.
Oreskes, Naomi, and Homer E. LeGrand. Plate Tectonics: An Insider’s History of the Modern Theory of the Earth. Boulder, CO.: Westview Press,
2003.
107. John S Wilkins john@wilkins.id.au
• Phylogenies used to be done qualitatively/
subjectively
19
Case: Phylogenies
From Stirton, R. A. 1940. Phylogeny of
North American Equidae. Bull. Dept.
Geol. Sci., Univ. California 25(4):
165-198.
108. John S Wilkins john@wilkins.id.au
• Phylogenies used to be done qualitatively/
subjectively
• In cases like the palaeontology of the horse, only
morphological-skeletal characters were available
19
Case: Phylogenies
From Stirton, R. A. 1940. Phylogeny of
North American Equidae. Bull. Dept.
Geol. Sci., Univ. California 25(4):
165-198.
109. John S Wilkins john@wilkins.id.au
• Phylogenies used to be done qualitatively/
subjectively
• In cases like the palaeontology of the horse, only
morphological-skeletal characters were available
• Consequently, the diagrams relied upon the
apprehension of small data sets combined with
the “expertise” of the taxonomist
19
Case: Phylogenies
From Stirton, R. A. 1940. Phylogeny of
North American Equidae. Bull. Dept.
Geol. Sci., Univ. California 25(4):
165-198.
110. John S Wilkins john@wilkins.id.au
• Phylogenies used to be done qualitatively/
subjectively
• In cases like the palaeontology of the horse, only
morphological-skeletal characters were available
• Consequently, the diagrams relied upon the
apprehension of small data sets combined with
the “expertise” of the taxonomist
• As phylogenetics became mathematised with
“numerical” taxonomy (computer-based, “total
evidence”), it was supposed to become more
objective
19
Case: Phylogenies
From Stirton, R. A. 1940. Phylogeny of
North American Equidae. Bull. Dept.
Geol. Sci., Univ. California 25(4):
165-198.
111. John S Wilkins john@wilkins.id.au
• Phylogenies used to be done qualitatively/
subjectively
• In cases like the palaeontology of the horse, only
morphological-skeletal characters were available
• Consequently, the diagrams relied upon the
apprehension of small data sets combined with
the “expertise” of the taxonomist
• As phylogenetics became mathematised with
“numerical” taxonomy (computer-based, “total
evidence”), it was supposed to become more
objective
• It wasn’t (choice of principal components skewed
the results unpredictably)
19
Case: Phylogenies
From Stirton, R. A. 1940. Phylogeny of
North American Equidae. Bull. Dept.
Geol. Sci., Univ. California 25(4):
165-198.
112. John S Wilkins john@wilkins.id.au
• Phylogenies used to be done qualitatively/
subjectively
• In cases like the palaeontology of the horse, only
morphological-skeletal characters were available
• Consequently, the diagrams relied upon the
apprehension of small data sets combined with
the “expertise” of the taxonomist
• As phylogenetics became mathematised with
“numerical” taxonomy (computer-based, “total
evidence”), it was supposed to become more
objective
• It wasn’t (choice of principal components skewed
the results unpredictably)
• Enter molecular data 19
Case: Phylogenies
From Stirton, R. A. 1940. Phylogeny of
North American Equidae. Bull. Dept.
Geol. Sci., Univ. California 25(4):
165-198.
113. John S Wilkins john@wilkins.id.au
• At first, DNA-DNA hybridisation
studies
20
Molecular phylogenies
Sibley, Charles G., and Jon E. Ahlquist. “Phylogeny and Classification
of Birds Based on the Data of DNA-DNA Hybridization.” Current
Ornithology. Boston, MA: Springer US, 1983.
114. John S Wilkins john@wilkins.id.au
• At first, DNA-DNA hybridisation
studies
• E.g., Sibley and Ahlquist’s
taxonomy of birds in 1980s
20
Molecular phylogenies
Sibley, Charles G., and Jon E. Ahlquist. “Phylogeny and Classification
of Birds Based on the Data of DNA-DNA Hybridization.” Current
Ornithology. Boston, MA: Springer US, 1983.
115. John S Wilkins john@wilkins.id.au
• At first, DNA-DNA hybridisation
studies
• E.g., Sibley and Ahlquist’s
taxonomy of birds in 1980s
• Based on percentage of
similarity of total DNA [T50H];
a single number representing
billions of bases
20
Molecular phylogenies
Sibley, Charles G., and Jon E. Ahlquist. “Phylogeny and Classification
of Birds Based on the Data of DNA-DNA Hybridization.” Current
Ornithology. Boston, MA: Springer US, 1983.
116. John S Wilkins john@wilkins.id.au
• At first, DNA-DNA hybridisation
studies
• E.g., Sibley and Ahlquist’s
taxonomy of birds in 1980s
• Based on percentage of
similarity of total DNA [T50H];
a single number representing
billions of bases
• “With few exceptions the
DNA comparisons “see” the
same clusters of species that
are seen by the human eye.”
20
Molecular phylogenies
Sibley, Charles G., and Jon E. Ahlquist. “Phylogeny and Classification
of Birds Based on the Data of DNA-DNA Hybridization.” Current
Ornithology. Boston, MA: Springer US, 1983.
117. John S Wilkins john@wilkins.id.au
• Full genomic sequences
21
Sequence data
118. John S Wilkins john@wilkins.id.au
• Full genomic sequences
• Molecular clocks
21
Sequence data
119. John S Wilkins john@wilkins.id.au
• Full genomic sequences
• Molecular clocks
• Problems
21
Sequence data
120. John S Wilkins john@wilkins.id.au
• Full genomic sequences
• Molecular clocks
• Problems
• Paralogy and
homology
21
Sequence data
121. John S Wilkins john@wilkins.id.au
• Full genomic sequences
• Molecular clocks
• Problems
• Paralogy and
homology
• Statistical
probabilities of
sequences
21
Sequence data
122. John S Wilkins john@wilkins.id.au
• Full genomic sequences
• Molecular clocks
• Problems
• Paralogy and
homology
• Statistical
probabilities of
sequences
• Hybridisation and
later transfer
21
Sequence data
123. John S Wilkins john@wilkins.id.au
• Full genomic sequences
• Molecular clocks
• Problems
• Paralogy and
homology
• Statistical
probabilities of
sequences
• Hybridisation and
later transfer
• What do the
diagrams summarise?
21
Sequence data
124. John S Wilkins john@wilkins.id.au
• A cladogram is a binary tree [B-Tree in computing]
22
Phylogenies are compressed data
Images: https://commons.wikimedia.org/wiki/File:Cladogram_stomiid.svg
http://fishesofaustralia.net.au/home/family/208
125. John S Wilkins john@wilkins.id.au
• A cladogram is a binary tree [B-Tree in computing]
• Its topology represents the analysed structure in the
very large data set
22
Phylogenies are compressed data
Images: https://commons.wikimedia.org/wiki/File:Cladogram_stomiid.svg
http://fishesofaustralia.net.au/home/family/208
126. John S Wilkins john@wilkins.id.au
• A cladogram is a binary tree [B-Tree in computing]
• Its topology represents the analysed structure in the
very large data set
• Contrary to Hollywood/TV, a biologist cannot
eyeball a DNA sequence and identify the
species
22
Phylogenies are compressed data
Images: https://commons.wikimedia.org/wiki/File:Cladogram_stomiid.svg
http://fishesofaustralia.net.au/home/family/208
127. John S Wilkins john@wilkins.id.au
• A cladogram is a binary tree [B-Tree in computing]
• Its topology represents the analysed structure in the
very large data set
• Contrary to Hollywood/TV, a biologist cannot
eyeball a DNA sequence and identify the
species
• In fact, they cannot even identify
alignments
22
Phylogenies are compressed data
Images: https://commons.wikimedia.org/wiki/File:Cladogram_stomiid.svg
http://fishesofaustralia.net.au/home/family/208
128. John S Wilkins john@wilkins.id.au
• A cladogram is a binary tree [B-Tree in computing]
• Its topology represents the analysed structure in the
very large data set
• Contrary to Hollywood/TV, a biologist cannot
eyeball a DNA sequence and identify the
species
• In fact, they cannot even identify
alignments
• The bigger the data set, the more compression of
the data is required to be able to understand the
relationships
22
Phylogenies are compressed data
Images: https://commons.wikimedia.org/wiki/File:Cladogram_stomiid.svg
http://fishesofaustralia.net.au/home/family/208
129. John S Wilkins john@wilkins.id.au
• A cladogram is a binary tree [B-Tree in computing]
• Its topology represents the analysed structure in the
very large data set
• Contrary to Hollywood/TV, a biologist cannot
eyeball a DNA sequence and identify the
species
• In fact, they cannot even identify
alignments
• The bigger the data set, the more compression of
the data is required to be able to understand the
relationships
• This is due to the fact that humans, and epistemic groups of
humans, have limitations on the working memory, cognitive
power, and information transmission channels, that they can bring
to a problem set, like any ML system
22
Phylogenies are compressed data
Images: https://commons.wikimedia.org/wiki/File:Cladogram_stomiid.svg
http://fishesofaustralia.net.au/home/family/208
130. John S Wilkins john@wilkins.id.au
The more complete a large data set, the more compression is required
to comprehend the data structure
23
Big data is not always better
131. John S Wilkins john@wilkins.id.au
The more complete a large data set, the more compression is required
to comprehend the data structure
A total census, for example, is reduced to trends, thresholds, averages,
etc. Economic activity is reduced to Gross Domestic Products, etc.
23
Big data is not always better
132. John S Wilkins john@wilkins.id.au
The more complete a large data set, the more compression is required
to comprehend the data structure
A total census, for example, is reduced to trends, thresholds, averages,
etc. Economic activity is reduced to Gross Domestic Products, etc.
In light of Big Data, even empirical adequacy becomes a statistical
property, as error creeping into measurement is magnified
23
Big data is not always better
133. John S Wilkins john@wilkins.id.au
The more complete a large data set, the more compression is required
to comprehend the data structure
A total census, for example, is reduced to trends, thresholds, averages,
etc. Economic activity is reduced to Gross Domestic Products, etc.
In light of Big Data, even empirical adequacy becomes a statistical
property, as error creeping into measurement is magnified
So, to understand, compress.
23
Big data is not always better
134. John S Wilkins john@wilkins.id.au
The more complete a large data set, the more compression is required
to comprehend the data structure
A total census, for example, is reduced to trends, thresholds, averages,
etc. Economic activity is reduced to Gross Domestic Products, etc.
In light of Big Data, even empirical adequacy becomes a statistical
property, as error creeping into measurement is magnified
So, to understand, compress.
Mind, compression is not, ipso facto, comprehension.
23
Big data is not always better
135. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
24
Conclusions
136. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
24
Conclusions
137. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
• Generalisation/-ability
24
Conclusions
138. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
• Generalisation/-ability
• Causal explanation
24
Conclusions
139. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
• Generalisation/-ability
• Causal explanation
which are pretty much de Regt’s three criteria restated.
24
Conclusions
140. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
• Generalisation/-ability
• Causal explanation
which are pretty much de Regt’s three criteria restated.
The initial epistemic activities of measurement, analysis and kinematic
generalisation all involve compression so that we may comprehend the things
needing explanation
24
Conclusions
141. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
• Generalisation/-ability
• Causal explanation
which are pretty much de Regt’s three criteria restated.
The initial epistemic activities of measurement, analysis and kinematic
generalisation all involve compression so that we may comprehend the things
needing explanation
• Compression of data is a kind of pattern matching, on the basis of which
ampliative inferences rest.
24
Conclusions
142. John S Wilkins john@wilkins.id.au
Science is, in my view, an attempt to express the greatest amount of empirical
consequences in the fewest terms. Understanding thus sets and constitutes:
• Empirical adequacy (van Fraassen)
• Generalisation/-ability
• Causal explanation
which are pretty much de Regt’s three criteria restated.
The initial epistemic activities of measurement, analysis and kinematic
generalisation all involve compression so that we may comprehend the things
needing explanation
• Compression of data is a kind of pattern matching, on the basis of which
ampliative inferences rest.
• Big data is not, in its own right, a good thing, so there!
24
Conclusions
144. John S Wilkins john@wilkins.id.au
25
For asking the question:
• Adam Ford
For discussion:
• The late John Collier
• Malte Ebach
Acknowledgements
145. John S Wilkins john@wilkins.id.au
25
For asking the question:
• Adam Ford
For discussion:
• The late John Collier
• Malte Ebach
• Ward Wheeler
• Marcus Hutter
Acknowledgements
146. John S Wilkins john@wilkins.id.au
25
For asking the question:
• Adam Ford
For discussion:
• The late John Collier
• Malte Ebach
• Ward Wheeler
• Marcus Hutter
For funding:
• ISHPSSSB Oslo conference organisers
For ear nibbles: Clio, a muse
Thank you
Acknowledgements