Author: Drew Wicke

Drew Wickehttp://drew.psib.org

Swarms

Innovation and learning seemed to develop more quickly with a larger population. Like comparing europe to the native americans. So I think that it would be wise to try to see how large groups of agents who were grouped doing different tasks whether innovation would emerge.

Like have a huge “soup” of agents and have different goals for different populations of agents. Will those with similar goals learn to cooperate.

If we limit resources will innovation emerge by having different populations cooperating to create a hirarchy to accomplish their respective goals?

Essentially, when a large population with differing goals but limited by the same resources will they learn to innovate in order to accomplish their goal or will they fail.

In the macro the micro complex actions apear to be simple effects. So, in essence we have a swarm.
Start listing of large multi agent systems that might also be swarms:

traffic networks including internet, cars, planes, and any other type of transportation routing.

learning in multi agent systems when others don’t learn?

Scope of Swarm

At what point does the macrophenoma of a MAS appear as a swarm like behavior?  Then in a large multiagent learning problem could the agents be viewed as swarms instead of agents?  Can an algorithm be created to discover the swarm behavior of a large system composed of MARL agents?

Adaptive websites

In AI terms the user agent’s environment when browsing the internet is their browser. So, if the browser could construct a model of the user’s behavior, when the user enters a website the model that has been constructed could be sent to the website in order to construct a personalized layout etc. Websites could possibly manipulate the user based on the provided model. I think that google already pretty much does this already with its search, but why stop there?

Usually when I go to a restaurant’s website I am looking for their menu. The browser could learn that is what I click when I go to those types of sites.

I could create a website that acts as a proxy to websites and could dynamically change the website for them! So, people go to my website to view other websites. I would create a plugin for chrome/firefox that would develop the profile when they search regular websites when they aren’t…

No, forget the website, just make the plugin! When they look at a website with the plugin running the website content will be rearranged and possibly home page will be displayed! Like in my personal example above when I go to a restaurant website it could show the contact info and menu!

This would mean that the user’s profile could be stored in the cloud or on there computer. If in the cloud I could compare the graphs of users visits and when a user goes to a website they have never been to, people that have that are similar to this person could help to reformat it! Sort of like Amazon or Netflix.

Of course the user could turn off the changing of the website and leave on the learning of the user.

We already have plugins to remove the ads why not do this? Is it possible to do this quick enough that the user doesn’t experience too much of a lag?

Culture

Could I use a model or description of the culture of an agent to determine how well it would adapt to working with others in a group.  Sort of a similarity index of cultures.  This would help an agent to decide if he wants to collaborate with a group.  Sean Luke has published two papers on how the use of culture has helped genetic programming (search culture http://www.cs.gmu.edu/~sean/papers/).

Skill Decomposition

In a cooperative multi-agent learning environment if a new agent enters the system the others can teach the new agent the high level task which it can then decompose to learn other useful skills.  In the single agent case an example would be teaching a robot to get from a to b which requires going through a door and the robot could learn the simpler activity of walking through a door by exploring its options space.  I think maybe task decomposition type algorithms could be augmented to decompose the option space.

Importing posts

Well I was able to export the wp_posts table from phpMyAdmin at freehostingcloud.  Then I had to change some of the id’s in the exported sql so that they didn’t conflict with those that were already used.  I imported the data into wp_posts.  Then I ran this sql

UPDATE wp_posts SET guid = replace(guid, ‘http://drew.freehostingcloud.com’,’http://drew.heliohost.org’);

so that the posts would actually work!

Old site

Hi,

I used to be at drew.freehostingcloud.com and I had to switch because for some reason freehostingcloud took my website and a bunch of other people’s sites on Sunday.  I even got on the forum to see what was up (http://www.freehostingcloud.com/forum/thread-can-t-access-website).  Thankfully I was able to still access the backend and get all of the data.  So, I switched to heliohost.  They are much faster, but occasionally I get a 500 error and I just have to refresh the page.