More Video Lectures

I've watched a few more video lectures recently, mostly focused around Complex Systems and Machine Learning. I haven't figured out the best process to internalise from this medium, although I suspect I will progress to something like Peteris Krumins' approach (for example, his analysis of Introduction to Algorithms). Anyway, some notes:

• Stochastic Search Methods by Bogdan Filipič (2005). Provides an introduction into search methods that use randomness, differentiated from calculus methods (like gradient decent) and enumerative methods (like exhaustive search and dynamic programming). The general field is motivated as approximation methods for knowledge discovery in high-dimensional and multi-modal search spaces (where calculus and enumerative methods break down). Focused introduction to Simulated Annealing and Evolutionary Algorithms including Evolution Strategies, Genetic Algorithm, and Genetic Programming.
  
• Dance Evolution by undergraduate students at the Evolutionary Complexity Lab, University of Central Florida (2007). Short demonstration of 3D dancers comprised of neural networks trained with an interactive genetic algorithm, specifically a variant of NEAT. More information and videos available on the project page.
• NERO 2.0: Neuro Evolving Robotic Operatives. An introduction to the nero game for interactively developing controllers for agent simulations using genetic algorithms and neural networks. I remember a keynote on this at a conference back in 2006 - looks like a lot of fun! Official page and project page.
• A Paradigmatic Complex System: The Immune System by Irun Cohen (2008). The presentation of the immune system as a model complex system, managing the subsuming complex system of the host organism. The immune system as computation, cytokine networks, and the system properties being degenerate, pleiotropic, and redundant. The increase in complexity can result in an increase in fragility (invertebrate versus vertebrate immune systems). There was also a great graph showing the trend of complexity against evolutionary time. I'm quite familiar with Cohen's take on the immune system, I researched it for a while for my dissertation, specifically the cognitive paradigm of immunity.
• Emergence of complexity in biological networks: from selection to tinkering by Ricard V. Solé (2008). Considers the pervasiveness of networks in complex systems, and considers the development of such networks under selection (optimization in some cases) and tinkering (working with what is available without foresight). Examples including protein interaction networks and word co-occurrences, each with model systems. The talk pointed me to an interesting paper entitled Evolution and Tinkering (1977) by François Jacob.
• Introduction to Complexity Science by Seth Bullock. Provides an introduction into complex systems with examples including termite mounds, the brain, developmental biology, and evolution. He provides an interesting case for the now pervasive use of evolution as an explanation and/or terminology out side of the field, a role that used to be taken by physics. A good rant at the end about biologically inspired techniques needing to be justified based on their innate suitability rather than their inspiration, tempered with the observation that we can still use tools that we do not fully understand (I ranted on this myself in my dissertation).
  

I have also been thinking about getting into some technical communication. I started with the notion of a book or an ebook providing an introduction to computational intelligence search techniques (something like Thierry Lecroq's book for String Algorithms and associated Java examples). Thinking that the presentation was a bit dry, my thoughts shifted to a blog or mini-ebook based approach (like Peepcode for ruby on rails development). I also think a screen cast medium would be effective for tutorials and demonstrations (like peepcode and railscasts). I want to find a good mix of interestingness, monetizability in a medium to present a set of discrete related topics.