What do you think?
Rate this book
Dynamics Of Complex Systems
The study of complex systems in a unified framework has become recognized in recent years as a new scientific discipline, the ultimate in the interdisciplinary fields. Breaking down the barriers between physics, chemistry, and biology and the so-called soft sciences of psychology, sociology, economics and anthropology, this text explores the universal physical and mathematical principles that govern the emergence of complex systems from simple components. Dynamics of Complex Systems is the first text describing the modern unified study of complex systems. It is designed for upper-undergraduate/beginning graduate level students, and covers a broad range of applications in a broad array of disciplines. A central goal of this text is to develop models and modeling techniques that are useful when applied to all complex systems. This is done by adopting both analytic tools, including statistical mechanics and stochastic dynamics, and computer simulation techniques, such as cellular automata and Monte Carlo. In four sets of paired, self-contained chapters, Yaneer Bar-Yam discusses complex systems in the context of neural networks, protein folding, living organisms, and finally, human civilization itself. He explores fundamental questions about the structure, dynamics, evolution, development and quantitative complexity that apply to all complex systems. In the first chapter, mathematical foundations such as iterative maps and chaos, probability theory and random walks, thermodynamics, information and computation theory, fractals and scaling, are reviewed to enable the text to be read by students and researchers with a variety of backgrounds.
864 pages, Paperback
First published September 10, 1997
10 people are currently reading
278 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 2 of 2 reviews
February 20, 2019
Looking at my notes, I've spent the last two months dedicating an hour per-day to reading Dynamics of Complex Systems by Yaneer Bar-Yam. Having went through it, was it worth it?
Let me preface the review with a short disclaimer: I will not be fair. Dynamics of Complex systems is a math book, one that demands either math skills I never had, or huge amounts of work. The review will be written from a perspective of a person that kinda sort-of gets on with maths and who read this book at a leisurely pace without putting much work in.
And from that perspective, I have some doubts about the utility of the book.
Complex systems are composites of nonisolable parts. This lack of isolation is what makes complex system complex: the whole cannot be understood by the usual method of decomposition. The book tries to tackle complex systems in two ways. First, by presenting tools and methods of analysis that are particularly useful in understanding complex systems. Second, by giving an in-depth analysis of several complex systems.
The first chapter of the book, Introduction and Preliminaries, introduces the simplest kind of complex system: the iterative map. What follows after that is a quick rundown through the methods of simulation and analysis that will be used throughout the book: thermodynamics and statistical mechanics, activated processed, cellular automata, statistical fields, Monte Carlo simulations, information theory, computation theory and fractals.
The rest of the book applies the above tools to understand some complex systems: artificial neural networks, the brain, protein folding, evolution, predator-pray models, modeling the complexity of animals and human culture.
For example, in the chapter about ANNs, the authors presents the mathematical model of simple feedforward and recursive networks, then shows how to understand them in the context of thermodynamics and statistical mechanics. Computer simulations are used to understand the capacity of Hopfield network. Subdivision, a characteristic of complex systems, is also examined in the context of the Hopfield network and this is segued into models of the mind and its phenomenology.
Each of the systems examined is fun to read about and has that "wow factor" when one thinks about it. Some chapters are more "mathy" than others. In particular, the chapters about human culture could be easily imagined in a popsci book, while the introductory chapters require serious mathematical chops.
And my biggest problems with the book concerns the approachability of the book. Multi-tome books could and have been written on each topic presented in the introductory chapter. Covering them all in what constitutes a third of the book makes the subject matter simply illegible. The author also lacks the gift of explaining things clearly. Even for subjects I was intimately familiar with (like neural networks or genetic algorithms) following along was a chore. Keeping up with methods of statistical mechanics was beyond me - I just tried to gloss over the equations and their transformations and try my best not to get overly lost.
I've been fascinated by system theory from quite some time while, but I always encounter the wall I've hit with this book. While, I find the generals concept mind-blowing, I have a hard time of finding take away lessons from the material. I doubt I could successfully apply the methods presented in the book.
Because of all of this it's hard for me to say if I'd recommend the book or not. The material is interesting but it's either written badly or for someone that's much smarter than me.
Let me preface the review with a short disclaimer: I will not be fair. Dynamics of Complex systems is a math book, one that demands either math skills I never had, or huge amounts of work. The review will be written from a perspective of a person that kinda sort-of gets on with maths and who read this book at a leisurely pace without putting much work in.
And from that perspective, I have some doubts about the utility of the book.
Complex systems are composites of nonisolable parts. This lack of isolation is what makes complex system complex: the whole cannot be understood by the usual method of decomposition. The book tries to tackle complex systems in two ways. First, by presenting tools and methods of analysis that are particularly useful in understanding complex systems. Second, by giving an in-depth analysis of several complex systems.
The first chapter of the book, Introduction and Preliminaries, introduces the simplest kind of complex system: the iterative map. What follows after that is a quick rundown through the methods of simulation and analysis that will be used throughout the book: thermodynamics and statistical mechanics, activated processed, cellular automata, statistical fields, Monte Carlo simulations, information theory, computation theory and fractals.
The rest of the book applies the above tools to understand some complex systems: artificial neural networks, the brain, protein folding, evolution, predator-pray models, modeling the complexity of animals and human culture.
For example, in the chapter about ANNs, the authors presents the mathematical model of simple feedforward and recursive networks, then shows how to understand them in the context of thermodynamics and statistical mechanics. Computer simulations are used to understand the capacity of Hopfield network. Subdivision, a characteristic of complex systems, is also examined in the context of the Hopfield network and this is segued into models of the mind and its phenomenology.
Each of the systems examined is fun to read about and has that "wow factor" when one thinks about it. Some chapters are more "mathy" than others. In particular, the chapters about human culture could be easily imagined in a popsci book, while the introductory chapters require serious mathematical chops.
And my biggest problems with the book concerns the approachability of the book. Multi-tome books could and have been written on each topic presented in the introductory chapter. Covering them all in what constitutes a third of the book makes the subject matter simply illegible. The author also lacks the gift of explaining things clearly. Even for subjects I was intimately familiar with (like neural networks or genetic algorithms) following along was a chore. Keeping up with methods of statistical mechanics was beyond me - I just tried to gloss over the equations and their transformations and try my best not to get overly lost.
I've been fascinated by system theory from quite some time while, but I always encounter the wall I've hit with this book. While, I find the generals concept mind-blowing, I have a hard time of finding take away lessons from the material. I doubt I could successfully apply the methods presented in the book.
Because of all of this it's hard for me to say if I'd recommend the book or not. The material is interesting but it's either written badly or for someone that's much smarter than me.
July 11, 2021
I started reading this book because I saw some works of Nassim Taleb cited it. So it should be fund to see what attracts the two intellectuals together.
(1) Main idea: Yaneer Bar-Yam explains complex systems. What are complex systems? Think of the dynamics within our body, the neuron's transmissions in the brain, and the nature. Taking the example of nature, we always have awe at how beautiful nature arranges all life forms, as well as its ability to recover from damage such as forest fires, or volcano eruption.
(2) Why scientific research in complex systems: The current science is all done in the form of simple systems. Think of random controlled trials. For example, when you want to argue that drinking more water will make you healthy, you need to have a twin who are very similar in their health condition and basic levels in the body. Then you make one person drink more water and the other drink less. If the one who drinks more become healthier after a month, then you can say that very possibly drinking more water does make a person healthier. We always repeat such a procedure when proving or disproving an argument, both in science and in life. But the book points out that there are limitations that such a simple system can never reach. For example, even we understand the a lot of properties of protein, cell, and electricity, we still struggle to understand what the neurons are doing in our brain. The core reason is that we need to understand the "interactions" among the neurons and all the single-point knowledge we have. Hence, the complex systems stands out as an individual field of science. Dynamics of Complex Systems describe some basic principles when treating the complex systems as a whole. The book makes examples of neural networks, protein folding, life, and society, each introduced by two chapters. If you are interested, you can dive into the topic that you like.
(3) Critical thinking: "Is complex systems research-able?" is the core question against the proposal of complex systems as a subject. I saw that even in this book, it constructs the field from very simple rules, such as the number series where x_{t+1} = f(x_t). Most conclusions about these complex systems are what we already know. Also, the bottomline is, if we are using the same research methodology as in single systems, this new field might not be tractable.
----
More readings:
- Excerpt "science can also focus on 'relations' rather than individual 'entities' ": https://necsi.edu/relational-properti...
- Excerpt "separation of scales": https://necsi.edu/11-separation-of-sc...
- Separation of time scales: http://www.bio-physics.at/wiki/index....
(1) Main idea: Yaneer Bar-Yam explains complex systems. What are complex systems? Think of the dynamics within our body, the neuron's transmissions in the brain, and the nature. Taking the example of nature, we always have awe at how beautiful nature arranges all life forms, as well as its ability to recover from damage such as forest fires, or volcano eruption.
(2) Why scientific research in complex systems: The current science is all done in the form of simple systems. Think of random controlled trials. For example, when you want to argue that drinking more water will make you healthy, you need to have a twin who are very similar in their health condition and basic levels in the body. Then you make one person drink more water and the other drink less. If the one who drinks more become healthier after a month, then you can say that very possibly drinking more water does make a person healthier. We always repeat such a procedure when proving or disproving an argument, both in science and in life. But the book points out that there are limitations that such a simple system can never reach. For example, even we understand the a lot of properties of protein, cell, and electricity, we still struggle to understand what the neurons are doing in our brain. The core reason is that we need to understand the "interactions" among the neurons and all the single-point knowledge we have. Hence, the complex systems stands out as an individual field of science. Dynamics of Complex Systems describe some basic principles when treating the complex systems as a whole. The book makes examples of neural networks, protein folding, life, and society, each introduced by two chapters. If you are interested, you can dive into the topic that you like.
(3) Critical thinking: "Is complex systems research-able?" is the core question against the proposal of complex systems as a subject. I saw that even in this book, it constructs the field from very simple rules, such as the number series where x_{t+1} = f(x_t). Most conclusions about these complex systems are what we already know. Also, the bottomline is, if we are using the same research methodology as in single systems, this new field might not be tractable.
----
More readings:
- Excerpt "science can also focus on 'relations' rather than individual 'entities' ": https://necsi.edu/relational-properti...
- Excerpt "separation of scales": https://necsi.edu/11-separation-of-sc...
- Separation of time scales: http://www.bio-physics.at/wiki/index....
Displaying 1 - 2 of 2 reviews