Kirsten Hacker posted: " My husband recently came home with a book that acts as a sort of Rosetta stone for data analysis methods in physics, chemistry, biology, and network science. He saw the author give a talk a few weeks ago and was impressed enough to buy a copy. https:"
My husband recently came home with a book that acts as a sort of Rosetta stone for data analysis methods in physics, chemistry, biology, and network science. He saw the author give a talk a few weeks ago and was impressed enough to buy a copy.
We had the book with us on our beach vacation in Denmark.
I've had pizza rolls before, but the Danish Pizza Snegle was superior in all respects: texture, flavor, cinnamon-roll consistency. Cinnabon should look into this concept.
While my husband would spend a whole day reading four pages of the book over and over while creating a simulation to demonstrate some concept, I read the whole thing in one day and spent my time imagining applications of the concepts to all sorts of systems while dreaming up ways to use those applications is fictional dramas. I would entitle the chapters along the lines of:
The Power Laws of Power
as when power distributes itself in a repeatable mathematical form in all sorts of systems
Self Organized Criticality
as when a pile of sand organizes itself into a critical shape and collapses in a discrete moment.
Punctuated Equilibrium
as when an system discretely evolves with episodes of stability interspersed with moments of change.
Seeking the Edge of Chaos
as when a system maximizes survival odds by perpetually resting between stability and chaotic change.
I was reminded that you can't narrow your perspective so much that you are blind to emerging possibilities and you can't sample random possibilities indefinitely. Optimal operation entails a balance and a definite selection.
But more than that, the book reminded me of the connections between a lot of jargon I've picked up during my education. For example --
A complex system that is path dependent and with memory is nonergodic and non-Markovian whereas a random walk is a Markov process.
Repeated convolution of random, spikey functions will produce a Gaussian distribution -- an attractor for iterated convolution in a recursive system.
For history or path dependent processes, the law of large numbers does not hold -- as when a coin is flipped and gets heads a million times in a row. This is why the intuition that more measurements always increases accuracy fails.
Chaos or diversity at one length scale increases when diversity is destroyed on another length scale as when positive entropy over one correlation length gives way to negative entropy over another.
After savoring the jargon, I settled on applications of the concepts to my problem with algorithms that deliver writing prompts to authors who want to plagiarize without feeling like they are plagiarizing. Primarily, the book gave me new ideas about how to plot data, but in the end it just seemed like it showed me 1000 ways to say the same thing and I wondered if keeping my data presentation simple is best. After all, being a show off with technical language only builds credibility in circles I rarely visit these days.
I prefer to stay grounded by applying the concepts to concrete examples like The Six Degrees of Kevin Bacon and building from there. If you aren't familiar with the game of The Six Degrees of Kevin Bacon, kids in the pre-internet era would test out their memory of film knowledge by trying to connect any actor to Kevin Bacon within 6 films. My dad was particularly good at this game because he liked to study the credits of every movie he watched and memorize the names.
Building from Kevin Bacon, it is easy to see how any network that has a certain amount of clustering and randomness will exhibit the property of being able to connect two elements together with a small number of steps. Given that it is easy for a human memory to sample a system with a set of six, I wonder if this feature is something intrinsic to nature or rather intrinsic to human pattern recognition capabilities.
I'm struck by how words tend to be arranged like social networks, such that meaning is preserved when the words are connected in a fractal pattern that mirrors what we see in nature while meaning is lost when they are connected randomly. After all, when our brains love seeing fractals and hate randomness they would love to see fractal clusters of words forming sentences, clusters of concepts forming paragraphs, and clusters of stories forming genres -- while links between genres extract broader understanding of the human condition. We love to see our inner mental state mirrored in the outer world -- or vice versa -- and our working memories tend to be limited to 6-14 elements in order.
This may seem trivial, but the number six has historically showed up in copyright standards and I wonder why the invention of the internet has led to the relaxation of this standard to fourteen. Has something changed about how words are clustered and how well we are able to extract meaning or about how well people are able to identify patterns and copying? People like stories that have a high degree of familiarity when their reading comprehension skills are low and this would lower their demand for novelty. Simultaneously, in a noisy internet environment, such a reader might also have an increased demand for repetition.
Yet how could the degrees separating words be growing in an age when the degrees separating people is dropping as a result of the internet? This would suggest that by growing more broadly connected socially we are losing our ability to communicate meaningfully as local connections are replaced with global connections.
more random connections and less clustering come to define our personal connections
fewer random, long range connections exist in language due to specialization and compartmentalization
sentences must become longer to express concepts that were once expressed in fewer words.
communication becomes less meaningful and efficient
people cannot share a common understanding of poetry
Given meme culture, this would be hard to deny that the internet is altering the structure of language itself, leading to a form of cultural Alzheimer's with tangled language becoming analogous to tangled proteins in a diseased, over-used brain.
When we wish to make sense, it helps to have a template or set of patterns that we can modify in ways that create new meaning without over-using our limited computing resources. A copying system that aids a human might be graphed out in steps of six in which the first five steps come from the same part of the copied story while the sixth moves the copying algorithm to the next part of the text to be mined.
I bet that my data of page number overlaps from books copied with machine assistance could be plotted in such a networked ring structure and show how a directly copied book illustrates a 'regular pattern', a book copied with the aid of a writing prompt generator would draws out a 'small-world pattern' and a book that was merely inspired by another would show a 'random pattern' as in the Watts-Strogatz diagrams below.
I like how this article connects this type of diagram to a 2-D grid that looks much like a labyrinth. It helps me connect this to concepts involving renormalization and percolation with which I'm familiar. I wonder if I might be able to turn the copying patterns of the books I've studied into such a 2-D grid and determine if a geometric fractal pattern can be identified.
If the book copying algorithm uses geometric fractals, perhaps a Sierpinski carpet will be observed if the page number overlaps are arranged in a grid. It is a wonderful way of visualizing how to square the circle.
My mind is still at the beach. That must be why my thoughts are jumping around so much. I want to go back. White noise is so soothing and most of the time, I felt washed clean of all thoughts. The eternal sunshine of the spotless mind.
Eh. Back to work, whatever that means. While on vacation, I sketched out (on paper!!!) five blog posts I'd like to write. They were just thoughts that seemed to require some sorting, resolving cognitive dissonance and relieving tensions wherever they appear. I think the motivation is to soothe myself and others.
Perhaps, I'm just drawn to the edge of chaos and need to explore possibilities in order to feel safe. When viewed from the seashore, I think that is a bad quality. But back in the city.. well. I miss the beach.
WordPress.com
No comments:
Post a Comment