Nearly caught up

OK, most of the things that were worrying me about not being done have been done. Oh, and camping was great fun. I haven't been camping in about 30 years. I honestly wasn't looking forward to it, but when I actually got out there and started hanging out with my son, it was a joy.

I missed my deadline for getting the satire done. But then again, I have made good progress with my paper on optimization techniques for my exemplar-based genetic algorithm fitness function.

I've also received my AmigoBots at work, so I've been doing research on how to do simultaneous localization and mapping (SLAM) using nothing but the (notoriously noisy) sonar range finders on the little guys.

Camping

I have written several of the papers that I need to, and now I am off to the wilds of Michigan to camp with my 8-year-old son, David for a few days.

What have I been up to?

I noticed the usual pattern in blog writing here: a period of time with one or more posts per day, followed by a week gap with the next couple of posts. The trick is not to slip into the "once a month with interspersed apologies" phase of blogging.

Unfortunately, this is one of the apologies. I've been swamped at work, writing research proposals. Actually, I only wrote a few this round, unlike the past where I have written as many as 12. I am learning new things about my ability to manage time in the midst of distractions.

Aside from the research proposals, I have been persuing an independent research project that has to do with Genetic Algorithms. I had originally thought that my idea for evolving Finite State Machines using genetic algorithms was really clever. Then I did the google search and discovered that, although clever, my idea was not original.

However, I did come up with a (hopefully!) novel way to improve the convergence of the genetic population to a solution. It involves using a dynamic fitness function, but one that is not based (directly, anyway) on the age of the population but rather on which problems in the problem set remain unsolved. The best way I have thought about expressing it to non-GA geeks so far is that it is like offering a prize for solving the hard problems. The amount of the prize is increased over time until the problem is finally solved. The effect of this is to cause the population to climb out of local maxima more quickly than can be accomplished using the usual mutation operators. Results that I hope to publish for next year's GECCO conference indicate a better than order of magnitude improvement in performance. So I have been working on the paper in MS Word and not on the blog.

And finally, I have papers to write for this years Simulation Interoperability Workshop. I am doing topics on using FSMs for simulation scenario authoring, using gestures of robots to anticipate their intentions, and using standard model interfaces to promote interoperability among military simulations. However, since all three papers are more than a week late, I may not get any of them published.

And really finally, I need to finish my satire for the WBA.

So maybe I'll write here instead of in word, so at least my (0) faithful readers can keep up with what I'm doing.

The key to success?

I saw a thing on T.V. (I was spending time with my wife, and t.v. is one of her "hobbies") the other day about a guy who traded a paperclip for a house by "trading up" 14 times. How does a person get away with something like that? So I asked him. Kyle, if you have anything to add, please respond here.

Another writing contest

The fine folks over at the WBA have another flash fiction carnival going on. This time, it is humor on the theme of writing and publishing. Watch this space for my entry.

Two Kinds of People

I spent a lot of time over the past few days re-installing operating systems and virus scanning my several household computers. This experience reminded me of two kinds of people:

There is a saying, "Mean people suck!" This is the type of person who writes computer viruses. I mean, I am a software developer and I like coding challenges, but I don't have the malicious streak that it takes to release a virus. So here's a giant raspberry for computer virus authors. pbtbtbtbtbt!

Then there are the folks who write free software. As I put the pieces of my system back together, I was gratified at how easy it was, largely because of the availability of free software that works. Software like Mozilla Firefox and Thunderbird and Cygwin. I don't share Richard Stallman's philosophy that all software MUST be free. But I am grateful for free software, much as a farmer is when his neighbors come over and help him raise a barn. So I tip my hat to the many authors of free software Tip! That's right neighborly of ya!

I will be giving back to the community by being a team coach for next year's RoboFest. More on that later.

Happy Independence Day

While you enjoy your hotdogs and fireworks, please don't forget to face East and salute our troops. I pray that they can come home soon.

Non-ja

I stumbled across some ninja crap on the internet today (ewwww, it smells). Here is the "official ninja home page". I was reminded of reading The Tick back in the late 80's. Here are some classic Tick quotes.

Another kind of flash...

Today, instead of playing WoW, I did the tutorials with Macromedia Flash MX. It is a very powerful tool, but it is perhaps too complicated. I am a research scientist, but I find its interface too complex to understand. On the other hand, I know a lot of artists who don't have any trouble with it. That should be a lesson on the meaning of "intelligence."