Neuronic Equations Revisited and Completely Solved
Since my first meeting with Cybernetics, it has been crystal clear to me that its true object is the study of “intelligence” (calling it “natural” or “artificial” is misleading and causes only confusion). Can a physicist, with his mind and tools, have a say in such matter? Although today the question easily receives brazenly affirmative answers (I fear that too many problems lay ahead still totally ignored), at the time of my first endeavours the situation stood quite differently. This was lucky, in that little bias could then poison a young, enquiring mind; luckier perhaps than today, when so many claim problems to be “almost solved” and pour down answers, when even questions cannot yet be soundly formulated. For this reason I shall refrain from sweeping statements and limit my discussion to only one of the three basic elements of my model (see below): a set of nonlinear equations that describe the behaviour of a system of coupled binary decision elements in discrete time (“neuronic equations”: NE). Their solution is in any case an essential preliminary to that of the remaining parts, “mnemonic equations” and “adiabatic learning hypothesis”, which are strongly connected to anatomical, or technological, structural information.
Unable to display preview. Download preview PDF.
- Caianiello ER (1974) Programma Procuste per l’analisi di linguaggi naturali. Calcolo 2:83 (1965);Google Scholar
- Caianiello ER Languages, hierarchical structures and logic. Coral Gables Conf “Orbis Scientiae”, Jan 1974Google Scholar
- Caianiello ER, Lauria F (1970) Il sistema nervoso centrale. Atti Conv Med Eur, Ist AngelisGoogle Scholar
- Caianiello ER, Capocelli R (1971) On form and language: The Procrustes algorithm for feature extraction. Kibernetik 8:233;Google Scholar
- Caianiello ER, Capocelli R Structural analysis of hierarchical systems. III Int Joint Conf Pattern Recognition, Nov 1976, CoronadoGoogle Scholar
- Caianiello ER, Grimson WEL (1975) Synthesis of boolean nets and time behaviour of a general mathematical neuron. Biol Cybern 18:111 (this formalism is connected to the classic results of S.W. Golomb, the so-called Rademaker-Walsh expansion, etc.)Google Scholar
- Mcculloch WS, Pitts W (1943) A logical calculus of ideas immanent in nervous activity. Bull Math Biophys 5Google Scholar
- Eigen M, Schuster P (1979) The hypercycle — principle of natural self-organization. Springer, Berlin Heidelberg New YorkGoogle Scholar