Abstract
Inspired by the extraordinary computing power promised by quantum computers, the quantum mind hypothesis postulated that quantum mechanical phenomena are the source of neuronal synchronization, which, in turn, might underlie consciousness. Here, I present an alternative inspired by a classical computing method with quantum power. This method relies on special distributed representations called hyperstrings. Hyperstrings are superpositions of up to an exponential number of strings, which—by a single-processor classical computer—can be evaluated in a transparallel fashion, that is, simultaneously as if only one string were concerned. Building on a neurally plausible model of human visual perceptual organization, in which hyperstrings are formal counterparts of transient neural assemblies, I postulate that synchronization in such assemblies is a manifestation of transparallel information processing. This accounts for the high combinatorial capacity and speed of human visual perceptual organization and strengthens ideas that self-organizing cognitive architecture bridges the gap between neurons and consciousness.
Similar content being viewed by others
Notes
PISA is available in the Online Resource. Its computing time is weakly-exponential (i.e., near-tractable), but here, the focus is on the special role of hyperstrings in it.
I ignore the common misconception that quantum computers also can solve the polynomial versus nondeterministic polynomial (P vs NP) problem of computing.
Formally, for functions \(f\) and \(g\) defined on the positive integers, \(f\) is \(O(g)\) if a constant \(C\) and a positive integer \(n_0\) exist such that \(f(n) \le C*g(n)\) for all \(n \ge n_0\). Informally, \(f\) then is said to be in the order of magnitude of \(g\).
The pencil selection example is close to the spaghetti metaphor in sorting (Dewdney 1984) but serves here primarily to illustrate that, in some cases, items can be gathered in one bin that can be dealt with as if it comprised only one item (hyperstrings are such bins).
For a given input string, the tree of hyperstrings and its hyperstrings are built on the fly, that is, the hyperstrings are transient in that they bind similar features in the current input only. This contrasts with standard PDP modeling, which assumes that one fixed network suffices for many different inputs.
References
Aerts D, Czachor M, De Moor B (2009) Geometric analogue of holographic reduced representation. J Math Psychol 53:389–398. doi:10.1016/j.jmp.2009.02.005
Ahissar M, Hochstein S (2004) The reverse hierarchy theory of visual perceptual learning. Trends Cogn Sci 8:457–464. doi:10.1016/j.tics.2004.08.011
Anderson JR (1983) The architecture of cognition. Harvard University Press, Cambridge
Atmanspacher H (2011) Quantum approaches to consciousness. In: Zalta EN (ed) The Stanford Encyclopedia of Philosophy. http://plato.stanford.edu/archives/sum2011/entries/qt-consciousness. Accessed 10 Jan 2015
Attneave F (1982) Prägnanz and soap-bubble systems: a theoretical exploration. In: Beck J (ed) Organization and representation in perception. Erlbaum, Hillsdale, pp 11–29
Bojak I, Liley DTJ (2007) Self-organized 40 Hz synchronization in a physiological theory of EEG. Neurocomputing 70:2085–2090. doi:10.1016/j.neucom.2006.10.087
Buzsáki G, Draguhn A (2004) Neuronal oscillations in cortical networks. Science 304:1926–1929. doi:10.1126/science.1099745
Buzsáki G (2006) Rhythms of the brain. Oxford University Press, New York
Byrne MD (2012) Unified theories of cognition. WIREs Cogn Sci 3:431–438. doi:10.1002/wcs.1180
Campbell SR, Wang DL, Jayaprakash C (1999) Synchrony and desynchrony in integrate-and-fire oscillators. Neural Comput 11:1595–1619. doi:10.1109/IJCNN.1998.685998
Chalmers DJ (1995) Facing up to the problem of consciousness. J Conscious Stud 2:200–219. doi:10.1093/acprof:oso/9780195311105.003.0001
Chalmers DJ (1997) The conscious mind: in search of a fundamental theory. Oxford University Press, Oxford
Churchland PS (1986) Neurophilosophy. MIT Press, Cambridge
Collard RF, Povel DJ (1982) Theory of serial pattern production: tree traversals. Psychol Rev 89:693–707. doi:10.1037/0033-295X.89.6.693
Cormen TH, Leiserson CE, Rivest RL (1994) Introduction to algorithms. MIT Press, Cambridge
de Barros JA, Suppes P (2009) Quantum mechanics, interference, and the brain. J Math Psychol 53:306–313. doi:10.1016/j.jmp.2009.03.005
Deutsch D, Jozsa R (1992) Rapid solutions of problems by quantum computation. Proc R Soc Lond A 439:553–558. doi:10.1098/rspa.1992.0167
Dewdney AK (1984) On the spaghetti computer and other analog gadgets for problem solving. Sci Am 250:19–26
Dijkstra EW (1959) A note on two problems in connexion with graphs. Numer Math 1:269–271. doi:10.1007/BF01386390
Dunn JM, Moss LS, Wang Z (2013) The third life of quantum logic: quantum logic inspired by quantum computing. J Philos Log 42:443–459. doi:10.1007/s10992-013-9273-7
Eckhorn R, Bauer R, Jordan W, Brosch M, Kruse W, Munk M, Reitboeck HJ (1988) Coherent oscillations: a mechanisms of feature linking in the visual cortex? Biol Cybern 60:121–130. doi:10.1007/BF00202899
Edelman GM (1987) Neural darwinism: the theory of neuronal group selection. Basic Books, New York
Felleman DJ, van Essen DC (1991) Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex 1:1–47. doi:10.1093/cercor/1.1.1
Feynman R (1982) Simulating physics with computers. Int J Theor Phys 21:467–488. doi:10.1007/BF02650179
Fingelkurts AA, Fingelkurts AA (2001) Operational architectonics of the human brain biopotential field: towards solving the mind-brain problem. Brain Mind 2:261–296. doi:10.1023/A:1014427822738
Fingelkurts AA, Fingelkurts AA (2004) Making complexity simpler: multivariability and metastability in the brain. Int J Neurosci 114:843–862. doi:10.1080/00207450490450046
Finkel LH, Yen S-C, Menschik ED (1998) Synchronization: the computational currency of cognition. In: ICANN 98 perspectives in neural computing, pp 23–40. doi:10.1007/978-1-4471-1599-1_3
Flynn MJ (1972) Some computer organizations and their effectiveness. IEEE Trans Comput C 21:948–960. doi:10.1109/TC.1972.5009071
Fries P (2005) A mechanism for cognitive dynamics: neuronal communication through neuronal coherence. Trends Cogn Sci 9:474–480. doi:10.1016/j.tics.2005.08.011
Fukushima K (1975) Cognitron: a self-organizing multilayered neural network. Biol Cybern 20:121–136. doi:10.1007/BF00342633
Gilbert CD (1992) Horizontal integration and cortical dynamics. Neuron 9:1–13. doi:10.1016/0896-6273(92)90215-Y
Gottesman D (1999) The Heisenberg representation of quantum computers. In: Corney SP, Delbourgo R, Jarvis PD (eds) Group22: Proc XXII Intern Colloq Group Theor Methods Phys. International Press, Cambridge, pp 32–43. arXiv:quant-ph/9807006
Gray CM, Singer W (1989) Stimulus-specific neuronal oscillations in orientation columns of cat visual cortex. Proc Natl Acad Sci USA 86:1698–1702. doi:10.1073/pnas.86.5.1698
Gray CM (1999) The temporal correlation hypothesis of visual feature integration: still alive and well. Neuron 24:31–47. doi:10.1016/S0896-6273(00)80820-X
Grover LK (1996) A fast quantum mechanical algorithm for database search. Proc 28th Ann ACM Symp Theor Comput, pp 212–219. arXiv:quant-ph/9605043
Gusfield D (1997) Algorithms on strings, trees, and sequences. Cambridge University Press, Cambridge
Hagar A (2011) Quantum computing. In: Zalta EN (ed) The Stanford Encyclopedia of Philosophy. http://plato.stanford.edu/archives/spr2011/entries/qt-quantcomp. Accessed 10 Jan 2015
Harary F (1994) Graph theory. Addison-Wessley, Reading
Hatfield GC, Epstein W (1985) The status of the minimum principle in the theoretical analysis of visual perception. Psychol Bull 97:155–186. doi:10.1037/0033-2909.97.2.155
Hebb DO (1949) The organization of behavior. Wiley, New York
Hinton GE (1990) Mapping part-whole hierarchies into connectionist networks. Artif Intel 46:47–75. doi:10.1016/0004-3702(90)90004-J
Hochberg JE, McAlister E (1953) A quantitative approach to figural “goodness”. J Exp Psychol 46:361–364. doi:10.1037/h0055809
Hochberg JE, Brooks V (1960) The psychophysics of form: reversible perspective drawings of spatial objects. Am J Psychol 73:337–354. doi:10.2307/1420172
Hochstein S, Ahissar M (2002) View from the top: hierarchies and reverse hierarchies in the visual system. Neuron 36:791–804. doi:10.1016/S0896-6273(02)01091-7
Hopcroft JE, Ullman JD (1979) Introduction to automata theory, languages, and computation. Addison-Wesley, Reading
Hubel DH, Wiesel TN (1968) Receptive fields and functional architecture of monkey striate cortex. J Physiol Lond 195:215–243
Kelso JAS (1995) Dynamic patterns: the self-organization of brain and behavior. MIT Press, Cambridge
Kimchi R (2003) Relative dominance of holistic and component properties in the perceptual organization of visual objects. In: Peterson MA, Rhodes G (eds) Perception of faces, objects, and scenes: analytic and holistic processes. Oxford University Press, New York, pp 235–263
Koffka K (1935) Principles of Gestalt psychology. Routledge and Kegan Paul, London
Köhler W (1920) Die physischen Gestalten in Ruhe und im stationären Zustand [Static and stationary physical shapes]. Vieweg, Braunschweig
Lamme VAF, Supèr H, Spekreijse H (1998) Feedforward, horizontal, and feedback processing in the visual cortex. Curr Opin Neurobiol 8:529–535. doi:10.1016/S0959-4388(98)80042-1
Lamme VAF, Roelfsema PR (2000) The distinct modes of vision offered by feedforward and recurrent processing. Trends Neurosci 23:571–579. doi:10.1016/S0166-2236(00)01657-X
Langley P, Laird JE, Rogers S (2009) Cognitive architectures: research issues and challenges. Cogn Syst Res 10:141–160. doi:10.1016/j.cogsys.2006.07.004
Leeuwenberg ELJ, van der Helm PA (1991) Unity and variety in visual form. Perception 20:595–622. doi:10.1068/p200595
Leeuwenberg ELJ, van der Helm PA, van Lier RJ (1994) From geons to structure: a note on object classification. Perception 23:505–515. doi:10.1068/p230505
Leeuwenberg ELJ, van der Helm PA (2013) Structural information theory: the simplicity of visual form. Cambridge University Press, Cambridge
Lehar S (2003) Gestalt isomorphism and the primacy of the subjective conscious experience: a Gestalt bubble model. Behav Brain Sci 26:375–444. doi:10.1017/S0140525X03000098
Marr D (2010) Vision. MIT Press, Cambridge (Original work published 1982 by Freeman)
Milner P (1974) A model for visual shape recognition. Psychol Rev 81:521–535. doi:10.1037/h0037149
Mourik V, Zuo K, Frolov SM, Plissard SR, Bakkers EPAM, Kouwenhoven LP (2012) Signatures of majorana fermions in hybrid superconductor-semiconductor nanowire devices. Science 336:1003–1007. doi:10.1126/science.1222360
Navon D (1977) Forest before trees: the precedence of global features in visual perception. Cogn Psychol 9:353–383. doi:10.1016/0010-0285(77)90012-3
Newell A (1990) Unified theories of cognition. Harvard University Press, Cambridge
Ozhigov Y (1999) Quantum computers speed up classical with probability zero. Chaos Solitons Fractals 10:1707–1714. doi:10.1016/S0960-0779(98)00226-4
Penrose R (1989) The emperor’s new mind: concerning computers, minds and the laws of physics. Oxford University Press, Oxford
Penrose R, Hameroff S (2011) Consciousness in the universe: Neuroscience, quantum space-time geometry and Orch OR theory. J Cosmol 14. http://journalofcosmology.com/Consciousness160.html. Accessed 10 Jan 2015
Plate T (1991) Holographic reduced representations: convolution algebra for compositional distributed representations. In: Proceedings of the 12th international joint conference on artificial intelligence, pp 30–35
Pollen DA (1999) On the neural correlates of visual perception. Cereb Cortex 9:4–19. doi:10.1093/cercor/9.1.4
Pomerantz JR, Sager LC, Stoever RJ (1977) Perception of wholes and their component parts: some configural superiority effects. J Exp Psychol Hum Percept Perform 3:422–435. doi:10.1037/0096-1523.3.3.422
Pylyshyn ZW (1999) Is vision continuous with cognition? The case of impenetrability of visual perception. Behav Brain Sci 22:341–423
Rinkus GJ (2012) Quantum computing via sparse distributed representations. NeuroQuantology 10:311–315. doi:10.14704/nq.2012.10.2.507
Rosenblatt F (1958) The perceptron: a probabilistic model for information storage and organization in the brain. Psychol Rev 65:386–408. doi:10.1037/h0042519
Rumelhart DE, McClelland JL (1982) An interactive activation model of context effects in letter perception: Part 2. The contextual enhancement effect and some tests and extensions of the model. Psychol Rev 89:60–94. doi:10.1037/0033-295X.89.1.60
Salinas E, Sejnowski TJ (2001) Correlated neuronal activity and the flow of neural information. Nat Rev Neurosci 2:539–550. doi:10.1038/35086012
Searle JR (1997) The mystery of consciousness. The New York Review of Books, New York
Seife C (2000) Cold numbers unmake the quantum mind. Science 287:791. doi:10.1126/science.287.5454.791
Sejnowski TJ, Paulsen O (2006) Network oscillations: emerging computational principles. J Neurosci 26:1673–1676. doi:10.1523/JNEUROSCI.3737-05d.2006
Shadlen MN, Movshon JA (1999) Synchrony unbound: a critical evaluation of the temporal binding hypothesis. Neuron 24:67–77. doi:10.1016/S0896-6273(00)80822-3
Shor PW (1994) Algorithms for quantum computation: discrete logarithms and factoring. In: Proceedings, 35th annual symposium on foundations of computer science, pp 124–134. doi:10.1109/SFCS.1994.365700
Stenger V (1992) The myth of quantum consciousness. Humanist 53:13–15
Sun R (2004) Desiderata for cognitive architectures. Philos Psychol 3:341–373. doi:10.1080/0951508042000286721
Suppes P, de Barros JA, Oas G (2012) Phase-oscillator computations as neural models of stimulus-response conditioning and response selection. J Math Psychol 56:95–117. doi:10.1016/j.jmp.2012.01.001
Tallon-Baudry C (2009) The roles of gamma-band oscillatory synchrony in human visual cognition. Front Biosci 14:321–332. doi:10.2741/3246
Tegmark M (2000) Importance of quantum decoherence in brain processes. Phys Rev E 61:4194–4206. doi:10.1103/PhysRevE.61.4194
Thagard P (2012) Cognitive architectures. In: Frankish K, Ramsay W (eds) The Cambridge handbook of cognitive science. Cambridge University Press, Cambridge, pp 50–70
Townsend JT, Nozawa G (1995) Spatio-temporal properties of elementary perception: an investigation of parallel, serial, and coactive theories. J Math Psychol 39:321–359. doi:10.1006/jmps.1995.1033
van der Helm PA (1988) Accessibility and simplicity of visual structures. PhD Dissertation, Radboud University Nijmegen
van der Helm PA, Leeuwenberg ELJ (1991) Accessibility, a criterion for regularity and hierarchy in visual pattern codes. J Math Psychol 35:151–213. doi:10.1016/0022-2496(91)90025-O
van der Helm PA, Leeuwenberg ELJ (1996) Goodness of visual regularities: a nontransformational approach. Psychol Rev 103:429–456. doi:10.1037/0033-295X.103.3.429
van der Helm PA, Leeuwenberg ELJ (1999) A better approach to goodness: reply to Wagemans (1999). Psychol Rev 106:622–630. doi:10.1037/0033-295X.106.3.622
van der Helm PA (2004) Transparallel processing by hyperstrings. Proc Natl Acad Sci USA 101:10862–10867. doi:10.107/pnas.0403402101
van der Helm PA, Leeuwenberg ELJ (2004) Holographic goodness is not that bad: reply to Olivers, Chater, and Watson (2004). Psychol Rev 111:261–273. doi:10.1037/0033-295X.111.1.261
van der Helm PA (2010) Weber–Fechner behaviour in symmetry perception? Atten Percept Psychophys 72:1854–1864. doi:10.3758/APP.72.7.1854
van der Helm PA (2012) Cognitive architecture of perceptual organization: from neurons to gnosons. Cogn Proc 13:13–40. doi:10.1007/s10339-011-0425-9
van der Helm PA (2014) Simplicity in vision: a multidisciplinary account of perceptual organization. Cambridge University Press, Cambridge
van Leeuwen C, Steyvers M, Nooter M (1997) Stability and intermittency in large-scale coupled oscillator models for perceptual segmentation. J Math Psychol 41:319–344. doi:10.1006/jmps.1997.1177
van Leeuwen C (2007) What needs to emerge to make you conscious? J Conscious Stud 14:115–136
van Rooij I (2008) The tractable cognition thesis. Cogn Sci 32:939–984. doi:10.1080/03640210801897856
VanRullen R, Thorpe SJ (2002) Surfing a spike wave down the ventral stream. Vis Res 42:2593–2615. doi:10.1016/S0042-6989(02)00298-5
Vassilieva E, Pinto G, de Barros JA, Suppes P (2011) Learning pattern recognition through quasi-synchronization of phase oscillators. IEEE Trans Neural Netw 22:84–95. doi:10.1109/TNN.2010.2086476
von der Malsburg C (1981) The correlation theory of brain function. Internal Report 81–2, Max-Planck-Institute for Biophysical Chemistry, Göttingen, Germany
Wagemans J, Feldman J, Gepshtein S, Kimchi R, Pomerantz JR, van der Helm PA, van Leeuwen C (2012) A century of Gestalt psychology in visual perception: II. Conceptual and theoretical foundations. Psychol Bull 138:1218–1252. doi:10.1037/a0029334
Wertheimer M (1912) Experimentelle Studien über das Sehen von Bewegung [On the perception of motion]. Z Psychol 12:161–265
Wertheimer M (1923) Untersuchungen zur Lehre von der Gestalt [On Gestalt theory]. Psychol Forsch 4:301–350
Wolfe JM (2007) Guided search 4.0: current progress with a model of visual search. In: Gray W (ed) Integrated models of cognitive systems. Oxford University Press, New York, pp 99–119
Acknowledgments
I am grateful to Julian Hochberg and Jaap van den Herik for valuable comments on earlier drafts. This research was supported by Methusalem Grant METH/08/02 awarded to Johan Wagemans (www.gestaltrevision.be).
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Appendix
Appendix
Structural information theory (SIT) adopts the simplicity principle, which holds that the simplest organization of a visual stimulus is the one most likely perceived by humans. To enable quantifiable predictions, SIT developed a formal coding model to determine simplest codes of symbol strings. This model provides coding rules for the extraction of regularities, and a metric of the complexity, or structural information load \(I\), of codes.
The coding rules serve the extraction of the transparent holographic regularities repetition (or iteration I), symmetry (S), and alternation (A). They can be applied to any substring of an input string, and a code of the input string consists of a string of symbols and coded substrings, such that decoding the code returns the input string. Formally, SIT’s coding language and complexity metric are defined as follows.
Definition 1
A code \(\overline{X}\) of a string \(X\) is a string \(t_1t_2 \ldots t_m\) of code terms \(t_i\) such that \(X = D(t_1)\ldots D(t_m)\), where the decoding function \(D : t \rightarrow D(t)\) takes one of the following forms:
for strings \(y\), \(p\), and \(x_i\) (\(i = 1,2,\ldots ,n\)). The code parts \((\overline{y})\), \((\overline{p})\), and \((\overline{x_i})\) are chunks. The chunk \((\overline{y})\) in an I-form or an A-form is a repeat, and the chunk \((\overline{p})\) in an S-form is a pivot which, as a limit case, may be empty. The chunk string \((\overline{x_1})(\overline{x_2})\ldots (\overline{x_n})\) in an S-form is an S-argument consisting of S-chunks \((\overline{x_i})\), and in an A-form, it is an A-argument consisting of A-chunks \((\overline{x_i})\).
Definition 2
Let \(\overline{X}\) be a code of string \(X=s_1 s_2\ldots s_N\). The complexity \(I\) of \(\overline{X}\) in structural information parameters (sip) is given by the sum of (a) the number of remaining symbols \(s_i\) (\(1 \le i \le N\)) and (b) the number of chunks \((\overline{y})\) in which \(y\) is neither a symbol nor an S-chunk.
The last part of Definition 2 may seem somewhat ad hoc, but has a solid theoretical basis in terms of degrees of freedom in the hierarchical organization described by a code. Furthermore, Definition 1 implies that a string may be encodable into many different codes. For instance, a code may involve not only recursive encodings of strings inside chunks—that is, from \((y)\) into \((\overline{y})\)—but also hierarchically recursive encodings of S- or A-arguments \((\overline{x_1})(\overline{x_2})\ldots (\overline{x_n})\) into \(\overline{(\overline{x_1})(\overline{x_2})\ldots (\overline{x_n})}\). The following sample of codes for one and the same string may give a gist of the abundance of coding possibilities:
Code 1 is a code with six code terms, namely, one S-form, two I-forms, and three symbols. Code 2 is an A-form with chunks containing strings that may be encoded as given in Code 3. Code 4 is an S-form with an empty pivot and illustrates that, in general, S-forms describe broken symmetry; mirror symmetry then is the limit case in which every S-chunk contains only one symbol. Code 5 gives a hierarchical recoding of the S-argument in Code 4. Code 6 is an I-form in which the repeat has been encoded into an A-form with an A-argument that has been recoded hierarchically into an S-form.
The computation of a simplest code for a string requires an exhaustive search for ISA-forms into which its substrings can be encoded, followed by the selection of a simplest code for the entire string. This also requires the hierarchical recoding of S- and A-arguments: A substring of length \(k\) can be encoded into \(O(2^k)\) S-forms and \(O(k2^k)\) A-forms, and to pinpoint a simplest one, simplest codes of the arguments of these S- and A-forms have to be determined as well—and so on, with \(O(\log N)\) recursion steps, because \(k/2\) is the maximal length of the argument of an S- or A-form into which a substring of length \(k\) can be encoded. Hence, recoding S- and A-arguments separately would require a superexponential \(O(2^{N \log N})\) total amount of work.
This combinatorial explosion can be nipped in the bud by gathering the arguments of S- and A-forms in distributed representations. The next definitions and proofs show that S-arguments and A-arguments group naturally into distributed representations consisting of one or more independent hyperstrings—which enable the hierarchical recoding of up to an exponential number of S- or A-arguments in a transparellel fashion, that is, simultaneously as if only one argument were concerned. Here, only A-forms \(\langle (y)\rangle /\langle (x_1)(x_2)...(x_n)\rangle \) with repeat \(y\) consisting of one element are considered, but Definition 4 and Theorem 1 below hold mutatis mutandis for other A-forms as well.
-
Graph-theoretical definition of hyperstrings:
Definition 3
A hyperstring is a simple semi-Hamiltonian directed acyclic graph \((V,E)\) with a labeling of the edges in \(E\) such that, for all vertices \(i,j,p,q \in V\):
where substring set \(\pi (v_1,v_2)\) is the set of label strings represented by the paths between vertices \(v_1\) and \(v_2\); the subgraph on the vertices and edges in these paths is a hypersubstring.
-
Definition of distributed representations called A-graphs, which represent all A-forms covering suffixes of strings, that is, all A-forms into which those suffixes can be encoded (see Fig. 9 for an example):
Definition 4
For a string \(T = s_1s_2\ldots s_N\), the A-graph \(\mathcal {A}(T)\) is a simple directed acyclic graph \((V,E)\) with \(V = \lbrace 1,2,\ldots ,N+1\rbrace \) and, for all \(1 \le i < j \le N\), edges \((i,j)\) and \((j,N+1)\) labeled with, respectively, the chunks \((s_i\ldots s_{j-1})\) and \((s_j\ldots s_N)\) if and only if \(s_i = s_j\).
-
Definition of diafixes, which are substrings centered around the midpoint of a string (this notion complements the known notions of prefixes and suffixes, and facilitates the explication of the subsequent definition of S-graphs):
Definition 5
A diafix of a string \(T = s_1s_2\ldots s_N\) is a substring \(s_{i+1}\ldots s_{N-i}\) (\(0 \le i < N/2\)).
-
Definition of distributed representations called S-graphs, which represent all S-forms covering diafixes of strings (see Fig. 10 for an example):
Definition 6
For a string \(T = s_1s_2\ldots s_N\), the S-graph \(\mathcal {S}(T)\) is a simple directed acyclic graph \((V,E)\) with \(V = \lbrace 1,2,\ldots ,\lfloor N/2 \rfloor + 2\rbrace \) and, for all \(1 \le i < j < \lfloor N/2 \rfloor + 2\), edges \((i,j)\) and \((j,\lfloor N/2 \rfloor + 2)\) labeled with, respectively, the chunk \((s_i\ldots s_{j-1})\) and the possibly empty chunk \((s_j\ldots s_{N-j+1})\) if and only if \(s_i\ldots s_{j-1} = s_{N-j+2}\ldots s_{N-i+1}\).
Theorem 1
The A-graph \(\mathcal {A}(T)\) for a string \(T = s_1s_2\ldots s_N\) consists of at most \(N+1\) disconnected vertices and at most \(\lfloor N/2 \rfloor \) independent subgraphs (i.e., subgraphs that share only the sink vertex \(N+1\)), each of which is a hyperstring.
Proof
-
(1)
By Definition 4, vertex \(i\) (\(i \le N\)) in \(\mathcal {A}(T)\) does not have incoming or outgoing edges if and only if \(s_i\) is a unique element in \(T\). Since \(T\) contains at most \(N\) unique elements, \(\mathcal {A}(T)\) contains at most \(N+1\) disconnected vertices, as required.
-
(2)
Let \(s_{i_1},s_{i_2},\ldots ,s_{i_n}\) (\(i_p < i_{p+1}\)) be a complete set of identical elements in \(T\). Then, by Definition 4, the vertices \(i_1,i_2,\ldots ,i_n\) in \(\mathcal {A}(T)\) are connected with each other and with vertex \(N+1\) but not with any other vertex. Hence, the subgraph on the vertices \(i_1,i_2,\ldots ,i_n,N+1\) forms an independent subgraph. For every complete set of identical elements in \(T\), \(n\) may be as small as \(2\), so that \(\mathcal {A}(T)\) contains at most \(\lfloor N/2 \rfloor \) independent subgraphs, as required.
-
(3)
The independent subgraphs must be semi-Hamiltonian to be hyperstrings. Now, let \(s_{i_1},s_{i_2},\ldots ,s_{i_n}\) (\(i_p < i_{p+1}\)) again be a complete set of identical elements in \(T\). Then, by Definition 4, \(\mathcal {A}(T)\) contains edges \((i_p,i_{p+1})\), \(p = 1,2,\ldots ,n-1\), and it contains edge \((i_n,N+1)\). Together, these edges form a Hamiltonian path through the independent subgraph on the vertices \(i_1,i_2,\ldots ,i_n,N+1\), as required.
-
(4)
The only thing left to prove is that the substring sets are pairwise either identical or disjoint. Now, for \(i < j\) and \(k \ge 1\), let substring sets \(\pi (i,i+k)\) and \(\pi (j,j+k)\) in \(\mathcal {A}(T)\) be not disjoint, that is, let them share at least one chunk string. Then, the substrings \(s_{i}\ldots s_{i+k-1}\) and \(s_{j}\ldots s_{j+k-1}\) of \(T\) are necessarily identical and, also necessarily, \(s_{i} = s_{i+k}\) and either \(s_{j} = s_{j+k}\) or \(j+k = N+1\). Hence, by Definition 4, these identical substrings of \(T\) yield, in \(\mathcal {A}(T)\), edges \((i,i+k)\) and \((j,j+k)\) labeled with the identical chunks \((s_{i}\ldots s_{i+k-1})\) and \((s_{j}\ldots s_{j+k-1})\), respectively. Furthermore, obviously, these identical substrings of \(T\) can be chunked into exactly the same strings of two or more identically beginning chunks. By Definition 4, all these chunks are represented in \(\mathcal {A}(T)\), so that each of these chunkings is represented not only by a path \((i,\ldots ,i+k)\) but also by a path \((j,\ldots ,j+k)\). This implies that the substring sets \(\pi (i,i+k)\) and \(\pi (j,j+k)\) are identical. The foregoing holds not only for the entire A-graph but, because of their independence, also for every independent subgraph. Hence, in sum, every independent subgraph is a hyperstring, as required.\(\square \)
Lemma 1
Let the strings \(c_1 = s_1s_2\ldots s_k\) and \(c_2 = s_1s_2\ldots s_p\) (\(k<p\)) be such that \(c_2\) can be written in the following two ways:
Then, \(X=Y\) if \(q=p/(p-k)\) is an integer; otherwise \(Y=VW\) and \(X=WV\), where \(V = s_1\ldots s_r\) and \(W = s_{r+1}\ldots s_{p-k}\), with \(r= p- \lfloor q \rfloor (p-k)\).
Proof
-
(1)
If \(1<q<2\), then \(c_2=c_1Wc_1\), so that \(Y=c_1W\) and \(X=Wc_1\). Then, too, \(r=k\), so that \(c_1=V\). Hence, \(Y=VW\) and \(X=WV\), as required.
-
(2)
If \(q=2\), then \(c_2=c_1c_1\). Hence, \(X=Y=c_1\), as required.
-
(3)
If \(q>2\), then the two copies of \(c_1\) in \(c_2\) overlap each other as follows:
$$\begin{aligned} \begin{array}{lccccccccccc} c_2 = c_1 X =s_1 &{} \ldots &{} s_{p-k} &{} s_{p-k+1} &{} \ldots &{} s_k &{} s_{k+1} &{} \ldots &{} s_p\\ c_2 = Y c_1 =&{} Y &{} &{} s_1 &{} \ldots &{} s_{2k-p} &{} s_{2k-p+1} &{} \ldots &{} s_k \end{array} \end{aligned}$$Hence, \(s_i=s_{p-k+i}\) for \(i=1,2,\ldots ,k\). That is, \(c_2\) is a prefix of an infinite repetition of \(Y\).
-
(3a)
If \(q\) is an integer, then \(c_2\) is a \(q\)-fold repetition of \(Y\), that is, \(c_2=YY\ldots Y\). This implies (because also \(c_2=Yc_1\)) that \(c_1\) is a \((q-1)\)-fold repetition of \(Y\), so, \(c_2\) can also be written as \(c_2=c_1Y\). This implies \(X=Y\), as required.
-
(3b)
If \(q\) is not an integer, then \(c_2\) is a \(\lfloor q \rfloor \)-fold repetition of \(Y\) plus a residual prefix \(V\) of \(Y\), that is, \(c_2=YY\ldots YV\). Now, \(Y=VW\), so that \(c_2\) can also be written as \(c_2=VWVW\ldots VWV\). This implies (because also \(c_2=Yc_1=VWc_1\)) that \(c_1=VW\ldots VWV\), that is, \(c_1\) is a \((\lfloor q \rfloor - 1)\)-fold repetition of \(Y = VW\) plus a residual part \(V\). This, in turn, implies that \(c_2\) can also be written as \(c_2=c_1WV\), so that \(X=WV\), as required.\(\square \)
Lemma 2
Let \(\mathcal {S}(T)\) be the S-graph for a string \(T = s_1s_2\ldots s_N\). Then:
-
(1)
If \(\mathcal {S}(T)\) contains edges \((i,i+k)\) and \((i,i+p)\), with \(k < p < \lfloor N/2 \rfloor + 2 - i\), then it also contains a path \((i+k,\ldots ,i+p)\).
-
(2)
If \(\mathcal {S}(T)\) contains edges \((i-k,i)\) and \((i-p,i)\), with \(k<p\) and \(i < \lfloor N/2 \rfloor + 2\), then it also contains a path \((i-p,\ldots ,i-k)\).
Proof
-
(1)
Edge \((i,i+k)\) represents S-chunk \((c_1) = (s_i\ldots s_{i+k-1})\), and edge \((i,i+p)\) represents S-chunk \((c_2) = (s_i\ldots s_{i+p-1})\). This implies that diafix \(D = s_i\ldots s_{N-i+1}\) of \(T\) can be written in two ways:
$$\begin{aligned} \begin{array}{ccccc} D &{} = &{} c_2 &{} \ldots &{} c_2\\ D &{} = &{} c_1 &{} \ldots &{} c_1 \end{array} \end{aligned}$$This implies that \(c_2\) (which is longer than \(c_1\)) can be written in two ways:
$$\begin{aligned} c_2 = c_1X \quad \hbox { with }\quad X = s_{i+k}\ldots s_{i+p-1}\\ c_2 = Yc_1 \quad \hbox { with }\quad Y = s_i\ldots s_{i+p-k-1} \end{aligned}$$Hence, by Lemma 1, either \(X = Y\) or \(Y = VW\) and \(X =WV\) for some \(V\) and \(W\). If \(X = Y\), then \(D = c_1Y\ldots Yc_1\) so that, by Definition 6, \(Y\) is an S-chunk represented by an edge that yields a path \((i+k,\ldots ,i+p)\) as required. If \(Y = VW\) and \(X =WV\), then \(D=c_1WV\ldots VWc_1\) so that, by Definition 6, \(W\) and \(V\) are S-chunks represented by subsequent edges that yield a path \((i+k,\ldots ,i+p)\) as required.
-
(2)
This time, edge \((i-k,i)\) represents S-chunk \((c_1) = (s_{i-k}\ldots s_{i-1})\), and edge \((i-p,i)\) represents S-chunk \((c_2) = (s_{i-p}\ldots s_{i-1})\). This implies that diafix \(D = s_{i-p}\ldots s_{N-i+p+1}\) of \(T\) can be written in two ways:
$$\begin{aligned} \begin{array}{ccccc} D &{} = &{} c_2 &{} \ldots &{} c_2\\ D &{} = &{} Yc_1 &{} \ldots &{} c_1X \end{array} \end{aligned}$$with \(X\)=\(s_{i-p+k}\ldots s_{i-1}\) and \(Y\)=\(s_{i-p}\ldots s_{i-k-1}\). Hence, as before, \(c_2 = c_1X\) and \(c_2 = Yc_1\), so that, by Lemma 1, either \(X=Y\) or \(Y = VW\) and \(X =WV\) for some \(V\) and \(W\). This implies either \(D=Yc_1\ldots c_1Y\) or \(D=VWc_1\ldots c_1WV\). Hence, this time, Definition 6 implies that both cases yield a path \((i-p,\ldots ,i-k)\), as required.\(\square \)
Lemma 3
In the S-graph \(\mathcal {S}(T)\) for a string \(T = s_1s_2\ldots s_N\), the substring sets \(\pi (v_1,v_2)\) (\(1 \le v_1 < v_2 < \lfloor N/2 \rfloor + 2\)) are pairwise identical or disjoint.
Proof
Let, for \(i < j\) and \(k \ge 1\), substring sets \(\pi (i,i+k)\) and \(\pi (j,j+k)\) in \(\mathcal {S}(T)\) be nondisjoint, that is, let them share at least one S-chunk string. Then, the substrings \(s_{i}\ldots s_{i+k-1}\) and \(s_{j}\ldots s_{j+k-1}\) in the left-hand half of \(T\) are necessarily identical to each other. Furthermore, by Definition 6, the substring in each chunk of these S-chunk strings is identical to its symmetrically positioned counterpart in the right-hand half of \(T\), so that also the substrings \(s_{N-i-k+2}\ldots s_{N-i+1}\) and \(s_{N-j-k+2}\ldots s_{N-j+1}\) in the right-hand half of \(T\) are identical to each other. Hence, the diafixes \(D_1=s_{i}\ldots s_{N-i+1}\) and \(D_2=s_{j}\ldots s_{N-j+1}\) can be written as
with \(p_1=s_{i+k}\ldots s_{N-i-k+1}\) and \(p_2=s_{j+k}\ldots s_{N-j-k+1}\). Now, by means of any S-chunk string \(C\) in \(\pi (i,i+k)\), diafix \(D_1\) can be encoded into the covering S-form \(S[C,(p_1)]\). If pivot \((p_1)\) is replaced by \((p_2)\), then one gets the covering S-form \(S[C,(p_2)]\) for diafix \(D_2\). This implies that any S-chunk string in \(\pi (i,i+k)\) is in \(\pi (j,j+k)\), and vice versa. Hence, nondisjoint substring sets \(\pi (i,i+k)\) and \(\pi (j,j+k)\) are identical, as required. \(\square \)
Theorem 2
The S-graph \(\mathcal {S}(T)\) for a string \(T = s_1s_2\ldots s_N\) consists of at most \(\lfloor N/2 \rfloor + 2\) disconnected vertices and at most \(\lfloor N/4 \rfloor \) independent subgraphs that, without the sink vertex \(\lfloor N/2 \rfloor + 2\) and its incoming pivot edges, form one disconnected hyperstring each.
Proof
From Definition 6, it is obvious that there may be disconnected vertices and that their number is at most \(\lfloor N/2 \rfloor + 2\), so let us turn to the more interesting part. If \(\mathcal {S}(T)\) contains one or more paths \((i,\ldots ,j)\) (\(i<j<\lfloor N/2 \rfloor +2\)) then, by Lemma 2, one of these paths visits every vertex \(v\) with \(i<v<j\) and \(v\) connected to \(i\) or \(j\). This implies that, without the pivot edges and apart from disconnected vertices, \(\mathcal {S}(T)\) consists of disconnected semi-Hamiltonian subgraphs. Obviously, the number of such subgraphs is at most \(\lfloor N/4 \rfloor \), and if these subgraphs are expanded to include the pivot edges, they form one independent subgraph each. More important, by Lemma 3, these disconnected semi-Hamiltonian subgraphs form one hyperstring each, as required. \(\square \)
Rights and permissions
About this article
Cite this article
van der Helm, P.A. Transparallel mind: classical computing with quantum power. Artif Intell Rev 44, 341–363 (2015). https://doi.org/10.1007/s10462-015-9429-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10462-015-9429-7