+ Ötletistálló +

Erre igazán büszkék vagyunk! close

+ MY-X DVD +

A MY-X team (fiatal tagjaival közösen) sikeresen vett részt a szarvasi konferencián!

A MY-X team (fiatal tagjaival közösen) sikeresen vett részt az Enyedi Emlékkonferencián!

A MY-X team egy tagja elnyerte a Bárdy Péter Természettudományi Díjat többek között kiemelkedő innovativitása okán!

A MY-X team egy tagjának a Gábor Dénes Középiskolai Ösztöndíj pályázatra benyújtott dolgozatát a Zsűri véleménye alapján az Alapítvány kuratóriuma matematikai különdíjban részesítette, ill. a kutatócsoport tagjának további tanulmányait egyszeri ösztöndíjjal segíti!

Különdíjat kapott a MY-X team egyik tagja a minszki nemzetközi esszéíró versenyen a jövő iskolájában várható IKT használatról szóló tanulmányáért!

III. helyezés a Hlavay Diákkonferencián!

Az NTP-NFTÖ támogatásával beszerzésre került egy prémium kategóriás laptop!

A TUDOK (nyíregyházi) országos döntőjéből sikeres továbbjutás a (székesfehérvári) Kárpát-medencei döntőbe!

Last modified: 2015.VII.19.14:08 - MIAÚ-RSS
A MIAÚ immár 20­+ éve áll a Köz szolgálatára!

Thought experiments I: Objectivity, sustainability, equilibrium

Leading article: 2009. May (MIAU No. 129.) - Pitlik László (MIAU) -
(Next article: MIAU No. 130.)

There were no leading articles in MIAU before. Useful documents that occurred monthly were offered without any commentaries or any kind of order. Of course, strategically important thoughts occurred among the articles, the WIKI-commentaries and the joint (MY-X Factor-Y) news, but those become guidelines only for attentive readers.
The introduction of leading articles to MIAU was considered for a long time, and it has become a necessity as the widespread usefulness of similarity analysis (which is the 'flagship' of the site even now) became more and more self-evident. The leading articles that are planned to be published monthly starting from the May of 2009 will try to put those universal tendencies in spotlight that, after reaching a critical mass (which relies heavily on the readers themselves) will determine our future in its foundations.
As how the mention of future is not only a rotund catchphrase, is proved by the following mosaic, the first thought experiment...

1/4 2/4 3/4 4/4

In medias res: The optimal taxation-system is like the bed of a fakir. It has many evil aspects, all frightening, we can still trust it thus it is sustainable.
The level of taxes in accordance with the democratic way can be determined relatively painlessly. Declare goals, make calculations about their costs then decide it in a democratic way if we want it for that much or not.
Of course in today's political atmosphere, even the mention of mutual interests may be regarded as a utopia. Thus it should be worthwhile to have some thoughts about the explanation of 'recognized necessity = happiness' principle on an objective and operative level.
If we compare the country's (hopefully valid) statistical databases with other countries' databases (which are hopefully valid too) on a regular basis (whereas the country is the object, and the induces are the attributes), then we get a learning pattern, which, if put into the correct spotlight (similarity analysis) will show how would the direction differ from the golden mean, that is observed unconsciously as a sustainable line by everyone.
We barely need anything else, than to ensure a real public potential of the nowadays often secure 'public' data, to organize millions of personal computers into a network, and the continuous presentation of the results (without embargo or any other kind of latency).
Quasi, at the emergence of each new data(row) facts may course through the continuously developing, robotized explaining algorithms, which draws a picture of strategic importance: what kind of pulling or pushing forces affect the objects. An object may be people, a company, a region, basically anything...
This draft marks a straight way to automatable strategy-making (for sci-fi enthusiasts, the Sheldon-plans), which is a key of legitimacy for each good governmental program, and the withdrawal of destructive demagoguery.
Of course, it should never be forgotten that human intuition drives everything, which is a productive force in case it lies in good will. However, if it is only for itself, then it bears the threats of Hocus-Pocus of Vonnegut, using the unconditional boldness of association as a source of power.
(Not incidentally arise the operationalism of the freedom and 'l'art pour l'art'-nature of speech and art, but it will be a center of attention in another article. Only to ruffle the temper: only the automatable science is useful, everything else is art, and in many cases, naive art, with all of its instinctively creative and heuristically popular aspects.)

Back to the fakir's bed (only one paragraph from now on!): The nails are the taxes, and we are to bear the contact with those formations of equal burden (c.f. taxes). If we organize everyone's (e.g. person, family, company) each (suggested by anybody) attribute into a common database, then proportional distribution of the planned tax collection can be 'degraded' to a mathematical question...
If someone consider the whole tax system a though haul, then we can reduce it to the planned property (assets) tax, in which case the counter- verification (by rational consideration, even the taxation) is question of similarity analysis, which do not differ from the aforementioned concept in any sense!
A less horrible metaphor than the fakir's bed: sustenance of balance is painful! Think of the preservation of body weight... Fattening is like economic growth: it feels good, but sooner or later it will lead to necessary corrections. And the diet is not simple. Like a constraint package... Even so if we did not mind that this little joy will have its drawback when we diverted from the equilibrium (c.f. Blow-Out). If someone diverts from the equilibrium, then he should know, whether he has to be happy, because it cannot be expected from the masses that they will learn to be happy with the correction stages. However, it is not impossible. Thus, balance seems to be pleasant sight only on strategic level or when looked back upon from the future. However, the constant monitoring of the loss of balance, and the constant counter-steering, mind-alteration is not simple...

And if the humane-oriented reader would endure this phalanster-suspicious? technocratic? frontal attack, then should be rewarded with a more humane chain of thoughts, the idealized view of learning:
We should open this new thought with the not-without-any intention marketing trick: learning (swotting) is not rewarding. It does not worth learning only to store knowledge. The world should be comprehended, instead of reflecting it. What must be done: The teacher pay attention to the environment and himself. The teacher learns about himself. The teacher teach/introduce himself to enquirers (C.f. Ayn Rand: Fountainhead, and the Dalai Lama). The learner get to know/understand the teacher. The learner understands himself. The learner (as an intuition-generator) makes himself useful. (What do you teach? I observe me! I learn me! I teach me! You learn me! You learn You! You use You!). Is he similarity analysis = case specific derivation that simple?! Of course the demagoguery-craft and corrupted price-performance analysis is not any more complicated, but it needs less cases, and the model can show already signals of (over)fitting. (I will further explain this in another time.)

The essay is a tough genre: it is seductively free and unlimited, unfortunately, however, the expert may see attack-incentive points with every written word, whereas the reader gets less and less confident. So, briefly:
How to prove the sobriety of the aforementioned thoughts? What is the proof itself?
One, who gets to deal with PhD-dissertations, will ask the question sooner or later: a dissertation is only like it was one, or did the author prove anything?
The proof in biometric approach is nothing else than the statement of significant differences of two treatments. It has its own canonized methodology, but in most cases life don't allow classic experiments with itself (e.g. weather forecast, stock exchange).
Otherwise (if we fix it) then each case (=starting state and its consequence) may be instructive. Is the margarine healthy, or choose butter instead? Who would not know those press releases that are referring to 'dubious' scientific results, and which are considered only to be marketing tricks...
If biometrics is not the universal method, then how to continue? How to distinguish a missionary/theological argument from a scientific proof? How to decide if consistency of a parliamentary speech is valuable, or demagogue based on its logic?
The answer lies in mathematical logic, which is hardly ever applied except some unique people... Even though we are referring to it continuously...
Our life is full of collections of cases that are not based on experiments, which we should understand instinctively, and what should help us to understand the effects of certain manipulations (c.f. production functions). Not because we didn't teach each other to do it consciously, but because since Konrad Lorenz, we know, that it can be dealt with only this way: intuition is based on instincts. The foundation of human knowledge (and the phenotypic knowledge of all beings) is inadvertent apprehension. Fortunately, people talk and write about things they suppose to know (heuristics). Sadly, however, the effectiveness of these two forms is clouded. Source code and mathematical formalism can be transferred without any loss, (without omission or overlap). And thus we got back to the line that distinguishes art from science. It is not enough to create something once, it shall be automated! And it is not enough to produce associations, they have to be proven!
And thus, the proof must be valid for data assets without experimental framework, for which the research methodology has no patents for. For example, does it worth inducting ERP-systems for companies? Shall we consider using this method in the following question: Is it worthwhile to develop exterior information systems?
The evaluation of questionnaires is often considered being a proven fact. But if we look back to the dawn of computer science, when Turing tests were to distinguish natural and artificial intelligence, the question was: Is it the computer or the people, who answer a certain question? Should not we ask ourselves: Are we able to distinguish, which answer(s) was generated by a random number generator, and which one was the valid, humane answer?
And if occur to us, that it is not so simple to decide, then we should not be so fast to consider the questions / the answers, and those who ask to be factual, like we do so in many cases nowadays... By the way, the similarity analysis can automatically decide who gave the most 'false' answers to certain questions, and to which questions were given the most 'humbug'?
It can be assumed that similarity analysis is some sort of strict proving procedure (viz. able to include all cases), that leads to the least inconsistent summary in multiple layers.
Contrary to that, cluster analysis seeks the mathematical possibility of classification without any direct goals, after the associations that are hard to distinguish from the suspicion of the Vonnegut-syndrome. There are some collections of cases, where it is not right to generalize, without running a similarity analysis, especially an Y0-model (viz. all forms of life and all the life spaces, because it can be proved that this kind of intuition generating, associative self-delusion does not make sense, thus there is no need to label them, (which is often called history-teaching and artistic impression). The legitimately subjective definition of disadvantage is a typical case where the environmental consciousness is trying to include everything in the objective function at last... However: biodiversity may not be optional in the future based on learning pattern of the past, which also means the mathematics of racism. Is it not curious? Until now, the material dual-faceted nature of the nuclear energy was the 'bogey man', and mathematics is going to show its true self, slowly but surely?!
Like everything, environmentalist movements that happen to be in the light of this positive frame are dual-faceted too: thus it is correct to stand against the more radical ('environ-mentalist') of those movements. For example, with the operationalism of the definition of sustainability, thus with this article, which have a simple message: if we cannot interpret each fact in a solid/robust model, then it is wrong to exclude these special facts!
For example, decision trees bears the same threats, that they seek valid rules only among those objects that are classified to the same nodes, but not system-wise.
In the model-based world the Vonnegut-like threat falls in the same category with the fear of overfitting, or technocratic demagoguery. Fortunately, multilayer models hold an effective shield over this: for example, in case of wine-price forecasts, a relatively good model may be created based on all available (more or less) independent variables. This, however, can be further refined based on the common errors and fittings of the price forecasts of diverse wine-types (in optional, but less and less effective iterations), but even a new point of view may be eligible to create the cornerstone of a framework, that is getting closer to perfection, but is still logical. And this means nothing else, than the identification of those learning error patterns, where the similarity analysis is able to learn the relation between the facts, and both the primary and the refined estimation. Forecasts will not be provided where it would prove to be wrong, instead of that there will be an 'I do not know' answer, so the expert system will self-restrict its own statements...
Closing words: is this the phalanster, or the way out? Natural selection will decide that! We are all exposed to so many problem-impulses here, in this little country, that existence precedes essence... so we have a chance to get intuition storms, which, if automated, will automate intuition itself (partially, but effectively). And if the mere existence of chess automatons did not yet urge someone to think where will the symbiosis of man-made machine and man lead, then it is indeed timely to do so now!

Please, send Your comments per email!