Automated selection of mathematical software

ABSTRACT

Current approaches to recommending mathematical software are qualitative and categorical. These approaches are unsatisfactory when the problem to be solved has features that can “trade-off” in the recommendation process. A quantitative system is proposed that permits tradeoffs and can be built and modified incrementally. This quantitative approach extends other knowledge-engineering techniques in its knowledge representation and aggregation facilities. The system is demonstrated on the domain of ordinary differential equation initial value problems. The results are significantly superior to an existing qualitative (decision tree) system.

REFERENCES

• 1: ADDISON, C. A. Implementing a stiff method based on the second derivative formulas. Dept. of Computer Science Tech. Rep. No. 130/79, Univ. of Toronto, 1979.
• 2: ADDISON, C. A., ENRIGHT, W. H., GAFFNEY, P. W, GLADWELL, I., AND HANSON, P. M. A decision tree for the numerical solution of boundary value ordinary differential equations. SMU Tecb. Rep. 89-7, Dept. of Mathematics, Southern Methodist University, Dallas, Texas, 1989.
• 3: C. A. Addison , W. H. Enright , P. W. Gaffney , I. Gladwell , P. M. Hanson, Algorithm 687: a decision tree for the numerical solution of initial value ordinary differential equations, ACM Transactions on Mathematical Software (TOMS), v.17 n.1, p.1-10, March 1991 [doi>10.1145/103147.103148]
• 4: AmD, T. J. The IMSL Library. In Sources and Development of Mathematical Software, W. R. Cowell, Ed., Prentice-Hall, Englewood Cliffs, N.J., 1984, pp. 264-301.
• 5: BATE, R., AND GYLYS, V. Intelligent workstations for scientists and engineers. Texas Instruments Computer Science Laboratory, manuscript, 1988.
• 6: BELL, D. E., KEENEY, R. L., AND RAIFFA, H., EDS. Conflicting Objectives in Decisions. Wiley, New York, 1977.
• 7: Ronald F. Boisvert , Sally E. Howe , David K. Kahaner, GAMS: a framework for the management of scientific software, ACM Transactions on Mathematical Software (TOMS), v.11 n.4, p.313-355, Dec. 1985 [doi>10.1145/6187.6188]
• 8: BUCHANAN, B. G., AND SHORTLIFFE, E. G. EDS. Rule-Based Expert Systems. Addison-Wesley, Reading, Mass., 1984.
• 9: BUZBEE, B.L. The SLATEC common mathematical hbrary. In Sources and Development of Mathematical Software, W. R. Cowell, Ed., Prentice-Hall, Englewood Cliffs, N.J., 1984, pp. 302-320.
• 10: DuBom, D. AND PRAD~, H. Po'~sibility Theory. Plenum Press, New York, 1988.
• 11: DUDA, R. O., HART, P. E., AND NmSSON, N.J. Subjective Bayesian methods for rule-based inference systems. In Proceedings of the 1976 National Computer Conference (New York, Jun. 1976), pp. 1075-1082.
• 12: DYKSEN, W. R., AND GRITTER, C. R. Expert systems for scientific computation and the algorithm selection problem. To appear in Expert Systems for Scientific Computing, E. Howtis, J. Rice, and R. Vichnevetsky, Eds., North-Holland, Amsterdam.
• 13: ERISMAN, A. H., NEVES, K. W., AND PHILIPS, I. R. The Boeing mathematical software library. In Sources and Development of Mathematical Software, W. R. Cowell, Ed., Prentice- Hall, Englewood Cliffs, N.J., 1984, pp. 321-345.
• 14: FELDMAN, S.I. The how and which of problem solving environments. In Problem Solving Environments for Scient~ftc Computing, B. Ford and F. Chatelin, Ed., Elsevier Science Publishers B. V , New York, 1987, pp. 23-32.
• 15: FODERARO, J K., SKOWLER, K. L., AND LAYER, K. The FRANZ LISP Manual, Univ. of California, Berkeley, 1983.
• 16: FORD, B., AND POOL, J. C. T. The evolving NAG library. In Sources and Development of Mathematical Software. W~ R. Cowell, Ed., Prentice-Hall, Englewood Cliffs, N J., 1984, pp. 375-397.
• 17: FORD, B., ^ND ILES, R. M. J The what and why of problem solwng environments. In Problem Solving Enwronments for Scientt~c Computing, B. Ford and F. Chatelin, Ed., Elsevier Science Publishers B. V., New York, 1987, pp. 3-18
• 18: P. A. Fox , A. P. Hall , N. L. Schryer, The PORT Mathematical Subroutine Library, ACM Transactions on Mathematical Software (TOMS), v.4 n.2, p.104-126, June 1978 [doi>10.1145/355780.355783]
• 19: P. W. Gaffney , J. W. Wooter , K. A. Kessel , W. R. McKinney, NITPACK: An Interactive Tree Package, ACM Transactions on Mathematical Software (TOMS), v.9 n.4, p.395-417, Dec. 1983 [doi>10.1145/356056.356058]
• 20: GAFFNEY, P. W. NEXUS: Toward a problem solving environment (PSE) for scientific computing. CMI Rep. CCS 86/2, Chr. Michelsen Inst., Dept. of Science and Technology, Fantoftvegen 38, N-5036 Fantoft, Bergen, Norway, 1986.
• 21: Ian Gladwell, Initial Value Routines in the NAG Library, ACM Transactions on Mathematical Software (TOMS), v.5 n.4, p.386-400, Dec. 1979 [doi>10.1145/355853.355856]
• 22: GLADWELL, I., AND LUCKS, M. An automated consultation system for the selection of mathematical software. In InteUigent Mathematical Software Systems, E. Houstis, J. Rice, and R. Vichnevetsky, Eds., North-Holland, Amsterdam, 1990, pp. 79-86.
• 23: GLADWELL, I. AND LUCKS, M. An interactive session with a knowledge based system for numerical software selection. To appear in the Transactions of the IMACS 13th World Congress, AI, Expert Systems and Symbolic Computing for Scientific Computatmn, E N. Houstis and J. R. Rice, Eds, North-Holland, Amsterdam, 1992.
• 24: GORDON, J, AND SHORTLIFFE, E.H. The Dempster-Shaffer theory of evidence. In Rule-Based Expert Systems, B. G. Buchanan and E. H. Shortliffe, Eds, Addison-Wesley, Reading, Mass., 1984, pp. 272-292.
• 25: Gopal K. Gupta , Ron Sacks-Davis , Peter E. Tescher, A review of recent developments in solving ODEs, ACM Computing Surveys (CSUR), v.17 n.1, p.5-47, March 1985 [doi>10.1145/4078.4079]
• 26: HALL, G., AND WATT, J. M Modern Numertcal Methods for Ordinary Differentml Equations. Clarendon Press, New York, 1976.
• 27: Harwell Subroutine Library. Computing Division, Harwe}l Laboratory, Ox~ordshire, England.
• 28: HAZEL, P. AND O'DoNoHOE, M.R. HELP numerical: The Cambridge interactive documentation system for numerical methods. In Production and Assessment of Numerical Software, M. L. Hennell and L. M Delves, Eds., Academic Press, London, pp. 367-382.
• 29: HWANC, C. L., AND MASUD, A. S.M. Multiple Objectwe Decision Making-Methods and Applicatzons: A State-of-the~Art Survey. Springer-Verlag~ New York, 1979
• 30: KAM~L, M., MA, K. S., AND ENRI~HT, W. H. ODEXPERT: An expert system to select numerical solvers for initial value ODE systems. To appear in Expert Systems for Scientific Computing, E. Houstis, J. Rice, and R. Vichnevetsky, Eds., North Holland, Amsterdam.
• 31: KEENEY, R L., AND RAIFFA, H. Decisions with Multiple Objectives: Preferences and Value Tradeoffs. Wiley, New York, 1976.
• 32: Michael Lucks, A knowledge-based framework for the selection of mathematical software, Southern Methodist University, Dallas, TX, 1990
• 33: LucKs, M., AND GLADWELL, I. A functional representation for software selection expertise. To appear in Expert Systems for Sc~entz~c Computing, E. Houstis, J. Rme, and R. Vichnevetsky, Eds., North-Holland, Amsterdam.
• 34: Constantin Virgil Negoita, Expert systems and fuzzy systems, Benjamin-Cummings Publishing Co., Inc., Redwood City, CA, 1984
• 35: RICE, J.R. The algorithm selection problem In Advances zn Computers, Vol. 15, Rubicoff and Yovits, Eds., Academic Press, San Diego, 1976, Calif., pp 65-118
• 36: SALLAS, W.M. Use of SEARCH in the IMSL libraries interactive documentation facilities IMSL Directmns 6, 3 (Third Quarter 1989), 2-4.
• 37: Klaus Schulze , Colin W. Cryer, NAXPERT: a prototype expert system for numerical software, SIAM Journal on Scientific and Statistical Computing, v.9 n.3, p.503-515, May 1988 [doi>10.1137/0909033]
• 38: SILVERT, W. Symmetmc summation: A class of operations on fuzzy sets. IEEE Trans. Syst. Man Cybern. SMC-9, 10 (1979), 657-659.
• 39: ZADEtt, L.A. The role of fuzzy logic in the management of uncertainty in expert systems. Fuzzy Sets and Syst. 11, (1983), 199-227.

INDEX TERMS

Primary Classification:
  G. Mathematics of Computing

Additional Classification:
  G. Mathematics of Computing
  G.1 NUMERICAL ANALYSIS
      G.1.7 Ordinary Differential Equations
          Subjects: Initial value problems

  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.4 Knowledge Representation Formalisms and Methods
          Subjects: Representation languages

General Terms:
Algorithms, Performance

Keywords:
ordinary differential equations, quantitative knowledge representation, software selection