How order can emerge from chaos : from a machine learning engineer perspective

Machine learning and chaos theory

Tuesday, January 21, 2020

As a machine learning practitioner and enthusiast for a while now, I have discovered some interesting facts about life and how a lot of what I used to describe as random is actually following well-known patterns and have somehow a predictable behavior.

I will not go deep into the technical details of machine learning techniques or how to implement them in this article but I will explain what leads me to review my definition of chaos and randomness and how in life order emerges from chaos.

“In my youth, I would have argued that life is just a series of random events devoid of any meaning. But as a data scientist, I have to recognize that sometimes patterns emerge. Undeniable patterns.” — Bertram Guilfoyle from Silicon valley HBO Tv Series

Machine learning

I did already define machine learning and explained its approaches in details in my article about artificial intelligence based approaches in malware detection . And the important thing to keep in mind is that machine learning is actually a set of algorithms that use statistical and mathematical methods in order to find patterns that underlie the numerical data it takes as input and formalize them in a set of mathematical equations called the model. The precision of these models and whether they explain well the data is related to a lot of factors and parameters like the quality of the data, the algorithms used, etc.

This changed a lot in human life already, discovering that a lot of events in our daily lives are linked together and that there are models describing these relations with certain error rate. These links and relations between two different events (variables) are called among mathematicians and statisticians correlations, they don’t mean causation, even though sometimes they might but they give us at least an idea on the straight of the relations between different events, and how to control them in the future or change them.

correlations

There are a lot of useful correlations in our daily life, a simple example is the relation between hospital nurse performance and a lot of other variables related to work environment, workload and the number of patients in emergency room in addition to time spent with the patient, this can help to know how many nurses to hire for example or how to organize nurses work shifts, etc. A related study by Lake ET, Sanders J, Duan R, Riman KA, Schoenauer KM, Chen Y shows that the nurse work environment warrants attention to promote health care quality, safety, and patient and clinician well-being.

The discovery of these relations raised some serious questions about the “engineering” that controls our lives: away from religious ideas, what we used to think about life was random events that, by chance, could come across those perfect combinations and could give somehow life and all the order we see in the universe. But with all the discoveries researchers make every day in different fields, this idea started to make no sense, because everything is related. Mathematically speaking, even the probability that all the complicated order and perfection we see could happen by chance is just impossible, especially in a short period of time that is the age of our universe. These new ideas give birth to a whole new field in science and philosophy called “complex systems and chaos theory”, that aims first to understand and properly define the term chaos, and then measure the complexity of systems controlling and guiding our lives.

the scientific term chaos refers to an underlying interconnectedness that exists in apparently random events.— J. Briggs, F.D. Peat

How can order emerge from randomness?

The physics professors J N Patterson Hume and Donald Ivey of the University of Toronto deployed their expertise — as well as some seriously elaborate sets — to demonstrate how, with enough data, highly predictable patterns can emerge from unpredictable events, and they could come up with interesting results as shown and the simulation tweeted by Roberto Alonso González Lezcano and as this film from 1956 by The Physical Science Study Committee (PSSC) shows.

The answer to the question is still not available. Scientifically and mathematically speaking the probability of such order to appear from random events is almost zero, that means maybe we still didn’t understand life and that there is a lot to explore in science, or what we probably need to re-think of what we define as random events, and that everything is controlled by hidden hands that did engineer everything.