• OK, it's on.
  • Please note that many, many Email Addresses used for spam, are not accepted at registration. Select a respectable Free email.
  • Done now. Domine miserere nobis.
Architect

Architect

Professional INTP
Joined
Last seen
Messages
6,691
Reaction score
0

Profile Posts Latest Activity Postings About

  • This is crazy: I had searched every post you've ever made with the keyword "sensor" and was about 60% of the way through building a lengthy and coherent argument with quotes and links to back up what I said, and then lost it all to the void.

    There is no way I'm doing it again. I had some real f'n momentum and it's gone.

    Sorry.
    Architect
    OK no problem.

    Apologies if I bristled but it's a hot button of mine when people take positions on me without opening a discussion first. On this point I think I come off as more sure about my position than I really am. MBTI is a statistical theory and I'm looking for trends.
    I will do then, just give me some time.

    The funny thing is that I had actually not read that thread (today's, that is), or even knew of its existence. I just dug that question out of thin air and threw it at you.

    I would like to try to find the time today to put together a coherent post to argue the point I would like to make.
    "If the clock at A synchronizes with the clock at B and also with the clock at C, the clocks at B and C also synchronize with each other." Perhaps a strange question to ask, but is this a transitive relation? Just read it and I came to think about transitive relations as I've encountered them with regards to identity, causality, indeed, it seems to creep up everywhere now.
    (Continuing on the joke)

    An excellent job. Go down to Luigi's on 3rd Avenue and ask for the number three, then mention Vinnie; you'll find your payment in a briefcase by the corner in ten minutes.

    -Duxwing
    (time to use that user title of yours for fun! :D )

    Milo's been giving physics some 'trouble' in the "Are there any philosophers that deal with time thread," so I'd likes ya to make 'im have a little 'accident,' capice?

    :D

    -Duxwing
    ProxyAmenRa
    As I initially thought, using the discrete Fourier transform technique required too much influence. I came up with a much simpler idea to deal with the non-stationary data. For each time step the change in the mean of the data is not great. The change in the variance is dependent on the interdependent variable. I thought I would use a weighted average of the last 20 time steps as an input variable. The results of this worked quite well. My model now has very high accuracy statistics for the training and validation sets and I avoid naive prediction.

    I will now do a three month forecast to determine the usefulness of the model.
    Meer
    Does your son still play Minecraft? There's this great mod called Computercraft that lets you make little robots that you program with Lua. Probably the most fun I've had with coding like, evar.
    Architect
    Checkout the Ask Architect thread for background - somebody wanted to see my real person picture. I'll be changing it back however.
    Architect
    Hi Cog,
    Sorry, I've become a minor celebrity and landed in something hot so am way overloaded, you'll have to develop your theory solo.
    Architect
    Then I'd take the FFT first (easy), then apply a Kalman filter (not so hard), and if I got really stuck make up a HMM.
    ProxyAmenRa
    The period is slightly constant. Whether or not this assumption will hold will become evident when I compare the model against validation data.

    Wish me luck.
    ProxyAmenRa
    I have some periodic non-stationary time series data which I am trying to model. It has non-linear relationships to the independent variables which explain it. In order to get the accuracy of the forecast model I require, I wish to use lags of the time series I am trying to forecast. Since the lags of the series are so highly correlated with the series itself, any modelling technique will weight the lag coefficients greater than the variables which explain the time series. This would result in naive forecasts. What I need to do is introduce a seasonality basis function into the model which will reduce the weighting of the lag coefficients. I wrote a Discrete Fourier Transform algorithm today which find the underlying trend of the data and treat deviations as noise.

    The problem I have in my mind about adding the basis function is that it does not actually have any casual relationship to the observed phenomena. I am wondering whether or not this is a methodologically sound approach.
    Architect
    I haven't done much in practice but have a little familiarity from classes. Go ahead and try me if you wish, I'd be curious to see what it is in any case.
    ProxyAmenRa
    Would you know much about statistical modelling? I have some questions on whether or not a method I wish to use is methodologically sound.
    Architect
    Your an INFJ? Makes sense, INFJ's go off into different directions that I find hard to follow sometimes. Yes it makes sense.
    own8ge
    I know I'm stupid, but does my argumentation at least make some sense? Because I always fear it doesn't, making me clarify endlessly.(Typical INFJ trait, but an INFJ trait is also to over estimate one's own ability. So yeah, paradoxes in my head constantly.)
    Cognisant
    Rest assured my thread is merely a ruse to trick them into accepting my nihilistic mechanistic universe, then their new god takes his mask off and mine is underneath.

    Could you imagine anything more horrifying? :D
    Architect
    I'm expecting things to come unraveled this year, but it's always hard to try and time it.
    Hehe, given that I already share a type with so many idiots it's not like I would feel insulted to have him join my club, but I don't see INTJ as a possibility at all.

    At least you're no longer seeing him as ES. :)
    Architect
    Sure, but generally I prefer these conversations to be public so everybody can benefit, unless there is a particular reason to keep it private?
    ElvenVeil
    Hello Architect :)

    I read many of your posts in the science/typology topics and I would love to discuss MBTI with you at one point :) comparing notes or what one would call such
    Okay, thanks. 'A million times better' than typology, huh? Sounds like it's either going to be gold or ridiculous!
    Architect
    Plenty of threads here about PodLair. A group of people who claim to have the next evolution in typology, but it's 'a million times better and so can't even be called typology'. They have goofy names and visuals, but their ideas are sound, at least what I've seen so far. Look around the forum, there are about a million threads including one I just started called "Pod'Lair review"
    Never heard of the Pod'Lair people. Clue me in?

    No problem regarding not trying to convince anybody. We just want to spread the word that we're open to arguments if people do want to make them. It's not like the typings at Celebrity Types are set in stone.

    Ben Stein: We can agree that he is by no means a typical INTP, at least! :) I'll ping you if/when we get around to writing up our rationale.
    Architect
    Thanks. And no problem on your decisions, I'm dubious of using MBTI to type others, especially public figures so aren't trying to convince anybody. The infamous Pod'Lair people seem to have an angle on that which I'm trying to break into.

    On Torvalds I'll take that back, you're probably correct. In person I think he brings out the Ne big time which threw me off. It shouldn't have as I Ne when out in the world.

    I'm sticking to Ben Stein however, he's out of the club :)
    There. :)

    Thanks for naming the other ones you think are misses.

    We did indeed read the entire thread and it's not that we don't see where you are coming from. We've had people question the Stein typing before and we're considering taking the time to write up our rationale to see what people think of it.

    Linus Torvalds is one we're quite sure about from watching videos. Of course we can't rule out that the man is different in person, but since we don't have the privilege of having access to him personally, we can only go by the material that is publicly available.

    We'd appreciate hearing your arguments regarding all the rest you mention (as well as arguments about Torvalds based on something we can check out for ourselves, of course!). Of course I can't promise that we will be convinced, but as you can see from Kasparov, we really are open to having things pointed out to us and we do change our minds sometimes. :)
    Architect
    No problem, and I don't need any credit - just credit "Archie" if you wish.

    Check out that thread, I think you have a number of misses. Ben Stein is absolutely not an INTP, look at the analysis on that page. Also ...

    Marie Curie - Kiersey is wrong, she is more likely an ISTJ
    Albert Speer - More of a promoter than a thinker, unlikely INTP.
    Linus Torvalds - I've met him, I doubt an INTP
    Sigourney Weaver - INFJ more likely
    Tina Fey - Don't know, but I'm doubtful she's an INTP
    Hi, I'm one of the admins of Celebrity Types. After reading your thread, we checked out Kasparov again, and we think you are absolutely right that we have mistyped him. Thank you for the correction. :) We've already taken him down from the INTP page, and we are going to move him to the INTJ page. We'd like to credit you on our News page, and so I'm writing to ask you for a name we may use for that purpose?
    Architect
    I told the story somewhere that I sold at the peak in 05 and have been renting since. House prices here are still at least 20% too high, we just moved into a higher end rental.
  • Loading…
  • Loading…
  • Loading…
Top Bottom