One of the quirks of our recent work lives is that I edit the project portfolio every 3 months (quarterly, to be specific). The Pragmatic Genetic Programming project is coming up near the top for next season’s list.

Actually, an edit of what’s there to make it flow a bit better is also in order.

On the plus side: I spent this last season doing typesetting and book design, so I’m a lot farther along towards finishing that book and the other couple than I had been! Take heart!

And thanks.

I really loved your articles in pragmatic gp site; they are completely practical and better than the many papers & books i read. Are you going to continue updating these articles ?

Well done.

Simon

http://ks12.net/portfolio/people-in-beta/

Greetings from Berlin,
Gabriel

Suppose there are 30 or 40 tasks for each page: capitalization, line breaks, end-of-page hyphenation, spelling, punctuation checks, structural document-level markup, blockquote identification, poetry, greek and hebrew transcription [sic], editing illustrations, mathematical typesetting, hyperlinking, table setting &c &c.

And assume that your scores depends on your history at being checked on those, and can go down if other people, later, edit something you’ve signed off on.

Spreading the work is only possible if you’re dealing with texts like The Campfire Girls Go Bananas.

Little-e entrepreneurship is the charming eccentricity that drives business innovation in our culture and economy. It’s a willingness to accept risks that others would shy away from, in exchange for eventual rewards nobody else can see. It’s the Earlie…

Dave

I saw your post on Cucumber. Have you been able to find anything like Cuke or RSpec that runs well in Python? Actually, I really need to get something going in IronPython if at all possible, so it would need to be “pure”-Python, with no C dll dependencies.

Thanks for your attention.
Best regards,
Dave

I just picked through some genetic algorithms, and I’m noticing some interesting trends.

- They tend to be good at identifying *roughly* where a local minimum is located in a space, but become terribly inefficient at progressing from there.

- Mutation and crossover rates need to be dynamically balanced based on the variance, especially in dynamic sets, to prevent premature homogenization of the population. There shouldn’t be a trigger for hyper-mutation, it should be dynamic.

- Non-mutated candidates are worthless except to add weight to the success of a trait. These candidates tend to be heavier than simply tracking a data point which can be used to determine ‘virility’ of a trait.

I was asked to copyedit a PhD dissertation written by a sweet, well-meaning grad student who could neither write a decent sentence nor think straight. She parroted Freire but when it came to her own course on “ethnicity,” she didn’t accept anything from her students save the answers SHE wanted. Conclusion of her thesis: her course was a success, because she thought it was, and because in the required end-of-the-year papers, her students told her what a great teacher she was and how much they had learned in her course.

Just how many high school students will write papers critical of the one with the power to grade?

Thanks for this post, it was one of those times when somebody else articulated really well what you’ve thought yourself for a while. Really I couldn’t agree more, apart from pointing out that many of us (particularly in the sciences) don’t exactly have the writing skills required to produce engaging academic writing. But still, we can but try (and we should).

Anyway, I’ve posted my thoughts on my blog here: http://bit.ly/clHDmz

Thanks again.

Alistair

This reminds me more than a little bit of what happens here. Where “here” is some mix of blogs, politics, new media world, and so forth. Of PZ Myers, anti-Bush blogging, dissing bad power law junk… stuff people have been doing in other contexts and being lauded as amusing.

Help me see the differences.

Is it the paper? Is it something about it having been printed on paper? Or the topic? Or what?

As for the aesthetic, I confess I find surface play with style, “difficulty”, etc. rather boring, and while he’s free to pursue those tastes, trying to shame me into thinking I am stupid and/or poor trash if I do not share them does not sit well with me.

Strangely, I do not get that feeling from it. I do, though, get a sense that diversity of literary theory has contracted over the last few decades. Maybe seven. Perhaps it’s because I’m comparatively limited to reading the most popularized critics, and they seem to more or less agree on matters of taste and preferences.

And I guess I’m in a position where being told I’m stupid makes some sense, as long as there’s some suggestion of what to do next attached to the report.

This is _not_ the review I ran across after my bookstore encounter, I can’t find that one again, but it’s pretty close to the one I did see.

In light of this, and your initial reactions, I doubt the work will go far. But in some useful way I find it makes an interesting apple (or, given your reaction, turd) rolled into the banquet. Life’s short. I’ll spend the time reading it, see what I can change it into, and buy the man his lunch. After all, so few of us have written so many published pages—even of crap. Let alone with footnotes.

This is not the review I ran across after my bookstore encounter, I can’t find that one again, but it’s pretty close to the one I did see.

I was wondering if there is an implementation of ParetoGP ? and any ideas of its performance compared to others such as SPEA2,NGSA-II

thank you!