classes at UFPE
Apr. 26th, 2003 08:34 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
guslClasses started this Tuesday.
I went to:
Grad Intro to AI
Grad ProgLan Paradigms
Undergrad/Grad Algorithms
Undergrad/Grad Musical Computing
Probably staying with only AI and Musical Computing. Algorithms is too elementary, and ProgLan Paradigms looks a lot like CS208 from Bucknell, although apparently deeper.
AI is taught by a very entertaining French actor. He encouraged our self-affirmation as RESEARCHERS, and not students; and he made fun of experimental physics papers (more pages listing authors than content). We will be using UML to do our ontologies, and the course seems very practically-oriented, like most courses here. (My economist uncle said: Brazil needs more technology transfer, not more theory. But he also said my comparative advantage IS theoretical work, so maybe Brazil is not the place for me. I agreed.)
In Music, we are first studying acoustics and psychoacoustics, and a lot of my old doubts about signal-processing are coming up again. I'm always adding the constraint of conservation of energy (or power) to see if things hold up. (Kinda similar to how I use information theory). So far, I haven't learned anything new, but it was only my first lesson.
I am such a theorist. In ProgLan paradigms, the prof showed us a list of required properties of a programming language: and one of them was "naturalness" for solving problems in some domain. So I asked: "How do you measure naturalness?". He said it was subjective. So I went on to ask if Kolmogorov (program-size) Complexity could provide an objective measure, but he didn't know.
There is a more theoretical course called "Semantics of Programming Languages", but it will be next semester.
I went to:
Grad Intro to AI
Grad ProgLan Paradigms
Undergrad/Grad Algorithms
Undergrad/Grad Musical Computing
Probably staying with only AI and Musical Computing. Algorithms is too elementary, and ProgLan Paradigms looks a lot like CS208 from Bucknell, although apparently deeper.
AI is taught by a very entertaining French actor. He encouraged our self-affirmation as RESEARCHERS, and not students; and he made fun of experimental physics papers (more pages listing authors than content). We will be using UML to do our ontologies, and the course seems very practically-oriented, like most courses here. (My economist uncle said: Brazil needs more technology transfer, not more theory. But he also said my comparative advantage IS theoretical work, so maybe Brazil is not the place for me. I agreed.)
In Music, we are first studying acoustics and psychoacoustics, and a lot of my old doubts about signal-processing are coming up again. I'm always adding the constraint of conservation of energy (or power) to see if things hold up. (Kinda similar to how I use information theory). So far, I haven't learned anything new, but it was only my first lesson.
I am such a theorist. In ProgLan paradigms, the prof showed us a list of required properties of a programming language: and one of them was "naturalness" for solving problems in some domain. So I asked: "How do you measure naturalness?". He said it was subjective. So I went on to ask if Kolmogorov (program-size) Complexity could provide an objective measure, but he didn't know.
There is a more theoretical course called "Semantics of Programming Languages", but it will be next semester.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
(no subject)
Date: 2003-04-26 04:53 pm (UTC)at...
Date: 2003-04-26 05:19 pm (UTC)Universidade Federal de Pernambuco!
(note how I follow hyperlinks ;) )
theoretician!
Date: 2003-04-26 05:24 pm (UTC)I'd measure 'naturalness' by polling domain experts while they're learning the new language.
For example, say we invent Fortran & we want to measure 'naturalness' for mechanical engineers. Take a bunch of mech engineers, show them the language, get them doing stuff with it, and survey them after the first day, first week, first month and first year!
The particular survey question doesn't much matter -- as long as it's consistent to compare across candidate languages for the problem domain.
- A Fellow Theorist
Re: theoretician!
Date: 2003-04-28 09:36 pm (UTC)I am actually interested an idealization of the programmer. Program size seems like a good objective measure, but I could probably find some faults with it if I were in a thinking mood.
Some people are interested in making computers understand and use natural language... but I'm more interested in making humans speak computer languages... formal languages can provide a certain precision, conciseness, etc., which are more fit for doing science... besides allowing us to talk to computers.
I am always struggling with the fact that humans are not wired to do logic naturally. But one day they will. (probably with real wires)
(no subject)
Date: 2003-04-27 10:09 am (UTC)(no subject)
Date: 2003-04-28 09:39 pm (UTC)