Generalization of an Elementary Perceiving and Memorizing Machine

The Elementary Perceiver and Memorizer (EPAM) is a context, and contains sufficient information to retrieve the model of human associative memory and the processes of image of the stimulus from the association memory when human verbal learning, written in the list-processing language necessary. IPL-V. A generalization of the EPAM model (called EPAM The complex image building processes also provide an assoIII), based on earlier experiments on EPAM 11, has been dative mechanism by means of which earlier learning can formulated and programmed. usefully be brought to bear on later learning. New stimuli arc In EPAM 111. stimulus configurations and their internal learned by associating together and modifying images of images are allowed indefinite complexity of sub-parts. Complex stimuli already learned. The model is an attempt to resolve internal images are built by associating together, in list structhe problem of how meaningful learning differs from nontures, tokens for retrieving information about the sub-parts meaningful learning, from the association memory. Empirical explorations of the model are concerned generally The concatenation of tokens by the image building process with variations in behaviour as a result of varying stimulus allows for the association of stimuli in many different contexts. and response meaningfulness. The token signifies the occurrence of a stimulus in a particular 3.ieM6HTapHan nporpaMMa jhh "H3yqaioinHx MauaeT coSbiTHe CTHMyjia b oco66hhom KOHTeKCTe, a npiiuihh" (3IIJIHM (EPAM)) npeacTaß^weTCH Moaejibio cyTCTByeT b aaHHOM naineKe aocTaToiHO HHdpopMauHH, accouHaTHßiioil nfiMfiTH h npoueccoß H3yiemm mob <rro6bi BooCpawenHe CTHMyjia moskho 6mjio Bbißec" y iciOBCKa; saiman Mo,Te.-ib narmcaHa b bh«6 si3uko, hs accouHaTHßHofl naMHTH b tom cnyqae, eoJiH sto A-1H pa3paf)OTKH CnHCKOB IPL 15. Ha OCHOBaHHH OKaHCeTCH Heo6xo«HMbiM. pe3y.ibTaTon Go.ice paiiHHx nonwTOK cihopMHyjinpoCjioiKHbie npoueccw, no3Bo^aiouiHe EPAM 111 CO3Ba.iacb ii cnporpaHMHpoßa,iacb ooo6meHHaH Moae^b naßaTb BoofipasKeHHH, Taicwe 3anaxaioT accoimaTHBEPAM jiß.inioumncHTaX Ha3bißaeMoft EPAM 111. Hiiit MexaHH3M, npn noMoiiw KOToporo npeAbiAymne B EPAM 111, mc.no noaiacTefl KOMiuieKCHbix <~thonuTH MoryT 3<h<beKTHßHbiM o6pa3oM BJiHHTb Ha cnemv.iob h BiiyTpeiiHHx hx n3oGpan<eHHH MomeT decayroiiuie. Hoßbie CTHMyjiu H3yiaioTCfl b pe3yjibTaTe npe.ie.-n.no yc.iovKHfiTbcu. OioiKHbie BiiyTpeHHHe accoiiHpoßaHHtf (tbm caMMM h Mo«H<bHKau.HH) HsyienH3oopn>KeHHn-iiMeiomHeMHoroypoßHeflHH(})opMauHH hux panbine BOo6pajKeHHft. MoaeJib npeacTaßJineT o no.tCTpyuTypectpohtch cnocoSoM accouHHpoßaHHH coCoft nonbrricy peuiHTb 3a«aqy o tom, KaKHM oopa3OM 3iia«iKon miMRKOB, npw noMomH kotopux cBe^eHHH pa3JinqaTb 3HaiaromHe CBeaenHH ot He3HaiaiomHx. o no.macTnx mo>kho BUBO«HTb H3 accouHaTHßHofl 3MnHpmiecKHe onwTbi npoßOflHmnecH Ha Mo^e^b, naMtiTii. Boo6me roßopn, hmciot c paaHUMH BH«aMH noße-3auen.ieHne iiaMeKOB npoueccoM co3«aHHH boosaeHHn, npoHCxofIHWHMH H3 pa3HOBHfIHOCTH 3HaienHfi pn>Kenmi no3Bo.ineT accou.HHpoßaTb cTHMy^u, npoHCTHMy^a h oTBeTa. cxoanmiio H3 Miioro pa3Hbix kohtckctob. HaMeK 03HaLa machine elementaire pour percevoir et mettre en mblir Information concernant les sous-parties, a partir de la moire (EPAM) est un modele de la memoire humaine d'assomemoire dissociation. ciation. et dcs process de la facon humaine d'apprendre L'enchainement dcs signes au moyen de l'image eUabor^e verbalement, 6crit en langage IPL-V. En partant dcs r«sulpermet d'associer dcs excitations dans un grand nombre de tnts de tests anteVieurs, on a dlabore' et programme" une contextes diffe>ents. Le signe montre qu'il existe une excitagcneralisation dv modele EPAM, appele EPAM 111. tion dans un contexte particulier et une information suffisante Sur EPAM 111, les configurations d'excitation et les images peut, grace ace signe, regene>er l'image de I'excitation a inteneures de ces configurations ont une complexitc indefinie partir de la memoire dissociation, quand cela est necessaire. de sous-parlies. Dcs images internes complexes sont dlaborees Les procedis complexes gui 61aborent l'image fournissent en groupant. en liste de structures, les signes destine* a retaaussi un m6canisme dissociation, permettant d'elaborer faci405 IX, 1] automata theory lement la toute premiere information a partir de la toute informations ayant un sens, de celles gui n'en ont pas. derniere. Les tests empiriques dv modele se rapportent gene>alement On decouvre de nouvelles excitations en groupant (et en aux variations de fonctionnement, celles-ci £tant le resultat modifiant) les images d'excitations deja connues. Cc modele d'excitations et de reponses variables. est un essai de r6soudre le probleme de la differentiation dcs