Construction of a smooth multivariate simulation environment from a finite one-to-one mapping

ABSTRACT

Complex computer-simulated environments are often sought in which autonomous agents can interact and their capabilities can be examined. Creating such simulated environments, however, can often be regarded as a very non-trivial undertaking. This work demonstrates how one can construct a smooth and continuous online environment which is easily computable and runs in polynomial time using a finite number of example function input/output pairs. Due to the manner of its construction, the newly created smooth mapping will give the same or similar outputs as the original one-to-one function it was derived from. Also, having no discontinuities, the constructed smooth mapping will not break should it be made to accept out-of-range or erroneous actions from an agent which engages it. Ultimately, this algorithm is used to construct an integral portion of a sequential environment which is utilized in a study to simulate the acquisition of phoneme sequence generating ability thought to occur in the brain.

REFERENCES

• 1: D. L. James and Risto Miikkulainen. Sardnet: A self organizing feature map for sequences. Advances in Neural Processing Systems, 1995.
• 2: T. Poggio and F. Girosi. Networks for approximation and learning. In Proceedings of the IEEE, volume 78 of 9, pages 1481--1497, 1990.
• 3: Reiner Schulz , James A. Reggia, Temporally asymmetric learning supports sequence processing in multi-winner self-organizing maps, Neural Computation, v.16 n.3, p.535-561, March 2004 [doi>10.1162/089976604772744901]
• 4: S. Singh. Distinctive Features: Theory and Validation. University Park Press, 1976.
• 5: S. Singh and J. Black. A study of twenty-six intervocalic consonants as spoken and recognized by four language groups. Journal of the Acoustic Society of America, 39(2):372--387, 1966.

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.6 SIMULATION AND MODELING
      I.6.8 Types of Simulation

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.0 General
          Subjects: Cognitive simulation

General Terms:
Algorithms

Keywords:
function construction, phoneme sequence acquisition