Complex Systems Modeling

1
Complex Systems Modeling at the
Instrumentation Physics Laboratory
Christopher Monterola
National Institute of Physics University of the
Philippines Diliman, Quezon City
2
Organization of Presentation I. The
Instrumentation Physics Laboratory (IPL) II.
Introduction to Complex Systems III. Published
Works Highlights A. Neural Networks B.
Cellular Automata C. Numerical Modelling IV.
Current Research
Guide Historical background and Inspiration
3
I. The Instrumentation Physics Laboratory
http//www.nip.upd.edu.ph/ipl
Tel. [phone] loc. 6312 or (63 2) 434
4232 Fax. [phone] e-mailipl_at_nip.upd.edu.p
h
4
II. Complex Systems A. Introduction Operational
Definition the whole is greater than the sum of
its part Examples Biological structures,
Granular materials, N-body problem Driving
Mechanism Self-organization Quantitative
Manifestation Power Law, Phase Transition
1 Open question Are there CS that are not
self-organized?
1 Per Bak, How Nature Works,
(Springer-Verlag New York, Inc., 1996).
5
II. Complex Systems B. Strategies B.1
Differential Equations Methods Runge
Kutta, Finite-Difference, Neural Networks 2
Emergence self-organized rules B.2 Cellular
Automata Emergence can be
self-organized or simple rules 3 2 C.
Monterola and C. Saloma, "Characterizing the
dynamics of constrained physical systems with
unsupervised neural network" Phys Rev E, 57,
1247R-1250R (1999). 3 S. Wolfram, A new kind
of science, (Wolfram Media, Inc, 2002).
6
III. Published Works Highlights A. Neural
networks in complex systems 1 E. Bagarinao
and C. Saloma, "Frequency analysis with Hopfield
encoding neurons," Phys Rev E 53, 5516-5521
(1996). 2 C. Monterola and C. Saloma,
"Characterizing the dynamics of constrained
physical systems with unsupervised neural
network" Phys Rev E, 57, 1247R-1250R (1998). 3
M. Quito, C. Monterola and C. Saloma, "Solving
N-body problems with neural networks," Phys. Rev.
Lett. 86 pp. 4741-4744 (2001). 4 C. Monterola
and C. Saloma, "Solving the nonlinear Schrodinger
equation with an unsupervised neural network,"
Opt. Express 9, 72-84 (2001). 5 C. Monterola,
M. Lim, J. Garcia and C. Saloma, "Feasibility of
a Neural Network as Classifier of Undecided
Respondents in a Public Opinion Survey," Int. J
Public Opinion Research 14, Issue 2, pp. 222-229
(2002). 6 C. Monterola, M. Lim, J. Garcia and
C. Saloma, "Accurate forecasting of the undecided
population in a public opinion poll," Journal of
Forecasting 21, pp 435-449 (2002). 7 C
Monterola C Saloma, "Noise-driven manifestation
of learning in mature neural networks" Physical
Review Letters 89, 188102 (2002).
7
III. Published Works Highlights A. Neural
networks in complex systems
Implementation
Hopfield Inspired Model
1 E. Bagarinao and C. Saloma, "Frequency
analysis with Hopfield encoding neurons,"
Phys Rev E 53, 5516-5521 (1996).
8
III. Published Works Highlights A. Neural
networks in complex systems
Approach Combine Lagrangian optimization in the
NN training
Data from R. Ripley, Modeling spatial patterns,
J.R.Stat.Soc. B39, 172 (1977).
2 C. Monterola and C. Saloma, "Characterizing
the dynamics of constrained physical systems
with unsupervised neural network" Phys Rev E,
57, 1247R-1250R (1998).
9
III. Published Works Highlights A. Neural
networks in complex systems
3 M. Quito, C. Monterola and C. Saloma,
"Solving N-body problems with neural networks,"
Phys. Rev. Lett. 86 pp. 4741-4744 (2001).
10
III. Published Works Highlights A. Neural
networks in complex systems
Sech pulse ? 1 ? 4
4 C. Monterola and C. Saloma, "Solving the
nonlinear Schrodinger equation with an
unsupervised neural network," Opt. Express 9,
72-84 (2001).
11
III. Published Works Highlights A. Neural
networks in complex systems
Answers of Respondents
272 Peripheral questions
1. Socio-economic and political issues 2.
Approval rating of other personalities/institution
s 3. socio demographic profile
5 C. Monterola, M. Lim, J. Garcia and C.
Saloma, "Feasibility of a Neural Network as
Classifier of Undecided Respondents in a Public
Opinion Survey," Int. J Public Opinion Research
14, Issue 2, pp. 222-229 (2002). 6 C.
Monterola, M. Lim, J. Garcia and C. Saloma,
"Accurate forecasting of the undecided
population in a public opinion poll," Journal of
Forecasting 21, pp 435-449 (2002).
12
III. Published Works Highlights A. Neural
networks in complex systems

• NN prediction is similar to that of the
  conventional method

5 C. Monterola, M. Lim, J. Garcia and C.
Saloma, "Feasibility of a Neural Network as
Classifier of Undecided Respondents in a Public
Opinion Survey," Int. J Public Opinion Research
14, Issue 2, pp. 222-229 (2002). 6 C.
Monterola, M. Lim, J. Garcia and C. Saloma,
"Accurate forecasting of the undecided
population in a public opinion poll," Journal of
Forecasting 21, pp 435-449 (2002).
13
III. Published Works Highlights A. Neural
networks in complex systems
A. Visual perception of cat R.M. Siegel, Physica
42 D, 385 (1990). B. Response of crayfish A.
Longtin, A. Bulsara, and F. Moss, Phys. Rev.
Lett. 67, 656 (1991). C. Hearing of squirrel
monkey M.C. Teich, S. Khanna and P. Guinney, J.
Stat. Phys. 70, 257 (1993). D. Speech
enhancement (Cochlear implants) for profoundly
deaf R.P. Morse and E.F. Evans, Nature Medicine
2, 928 (1996). E. Location of prey by cricket J.
Levin and J.P. Miller, Nature 380, 165 (1996).
7 C Monterola C Saloma, "Noise-driven
manifestation of learning in mature neural
networks" Physical Review Letters 89, 188102
(2002).
14
III. Published Works Highlights A. Neural
networks in complex systems
F. Visual perception of human beings E.
Simonotto, et al., Phys. Rev. Lett. 78, 1186
(1997). G. Paddle fish feeding D. Russel, L.
Wilkens and F. Moss, Nature 402, pp.
291(1999). P. Greenwood et al, Phys. Rev. Lett.
84 , 4773 (2000). H. Rat skin sensing D. Nozaki,
D. Mar, P. Grigg J. Colins, Phys. Rev. Lett.
82, 2402 (1999). I. Blood regulation in human
beings I. Hidaka, D. Nozakiand Y. Yamamoto, Phys.
Rev. Lett. 85, 3740 (2000).
7 C Monterola C Saloma, "Noise-driven
manifestation of learning in mature neural
networks" Physical Review Letters 89, 188102
(2002).
15
III. Published Works Highlights A. Neural
networks in complex systems

• Image frame rate 60 Hz (16.6 ms / frame)
• Ath Perceptive contrast threshold
• Noise generation Gaussian/ temporal
• 256 x 256 pixels

E. Simonotto, et al., Phys. Rev. Lett. 78, 1186
(1997).
16
III. Published Works Highlights A. Neural
networks in complex systems
A 0.4
(a) multifrequency (b) bandpass (c) low pass (d)
high pass
7 C Monterola C Saloma, "Noise-driven
manifestation of learning in mature neural
networks" Physical Review Letters 89, 188102
(2002).
17
III. Published Works Highlights B.
Cellular Automata in complex systems
8 M. Lim and C. Saloma, "Emergence of
hysteresis in a network of nonhysteretic agents
with continuous responses," Phys. Rev. Lett. 88,
038701 (2002). 9 G.J. Perez, G. Tapang, M.
Lim and C. Saloma, "Streaming, disruptive
interference and power-law behavior in the exit
dynamics of confined pedestrians," Physica A
3123-4, pp 609-618 (2002). 10 D.E. Juanico, C.
Monterola C. Saloma, "Allelomimesis as generic
clustering mechanism for interacting agents,"
Physica A 320C, pp 590 - 600 (2003).
18
III. Published Works Highlights B.
Cellular Automata in complex systems
FIG. 4. (a) Oxygen isotope variation (d18OO) with
ice core depth that exhibits the YD cold event
(Greenland Ice Core Project), (b) Model, and (c)
network input signal.
FIG. 1. Network response.
8 M. Lim and C. Saloma, "Emergence of
hysteresis in a network of nonhysteretic agents
with continuous responses," Phys. Rev. Lett.
88, 038701 (2002).
19
III. Published Works Highlights B.
Cellular Automata in complex systems
9 G.J. Perez, G. Tapang, M. Lim and C. Saloma,
"Streaming, disruptive interference and
power-law behavior in the exit dynamics of
confined pedestrians," Physica A 312/3-4, pp
609-618 (2002).
20
III. Published Works Highlights B.
Cellular Automata in complex systems
System ? exptl ? ?num Sardinella schools
0.95 0.997 1.05 0.03 African buffalo
herds 1.15 0.995 1.20 0.06 Free-swimming
tuna 1.49 0.99 1.34 0.03 US financial
firms 2.06 0.7 2.05 0.15 Towns around Berlin
2.06 0.7 2.05 0.15 Barrios in
Cebu 2.18 0.6 2.20 0.15
Explain Power Law Demonstrate Phase Transition
10 D.E. Juanico, C. Monterola C. Saloma,
"Allelomimesis as generic clustering mechanism
for interacting agents," Physica A 320C, pp 590 -
600 (2003).
21
III. Published Works Highlights C.
Numerical Modeling in complex systems
11 F. Domingo and C. Saloma, "Phase transitions
in the confined growth of chains formed by
self-avoiding hard spheres," Physica D, Vol 141,
pp. 80 - 90 (2000).
12 J Bantang, M Lim, C Monterola and C Saloma,
"Gravity-assisted segregation of granular
materials of equal mass and size," Physical
Review E 66, 041306 (2002).
22
III. Published Works Highlights C.
Numerical Modeling in complex systems
11 F. Domingo and C. Saloma, "Phase transitions
in the confined growth of chains formed by
self-avoiding hard spheres," Physica D, Vol 141,
pp. 80 - 90 (2000).
23
III. Published Works Highlights C.
Numerical Modeling in complex systems
12 J Bantang, M Lim, C Monterola and C Saloma,
"Gravity-assisted segregation of granular
materials of equal mass and size," Physical
Review E 66, 041306 (2002).
24
IV. Current Research in complex systems
M. Quito - Travelling ant in the rain model
Gossip Spreading J. Romero C. Samson - Spatial
Distribution of water droplets J. Soriano -
Urinal problem M. Honrado - Network
robustness J. Bantang - Theory of cellular
automata Granular mixing P. Castro - Forex
Dynamics T. Pulido - Economic systems M.
Litong - Clogging/Filter problem M. Marfil -
IKOT jeep modeling T. Travers - protein
structure
25
Acknowledgements
Http www/members.fortunecity.com/camarila/fantasy
.html
Http www.novelart.com/