Data processing: artificial intelligence – Machine learning – Genetic algorithm and genetic programming system
Reexamination Certificate
1997-10-14
2004-05-25
Davis, George B. (Department: 2121)
Data processing: artificial intelligence
Machine learning
Genetic algorithm and genetic programming system
C706S047000
Reexamination Certificate
active
06741973
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to a method of generating a simulation model of consumers in a transaction environment, such as the behavior of customers in a bank branch.
Conventionally, a computer-based model of the behavior of consumers in a transaction environment is built by an expert using a combination of education (incorporating at least part of the previous model) and experience (altering a previous model or creating a new model). The expert selects and programs-in the characteristics of consumer behavior.
As transaction business, especially financial transaction business, becomes more complex, such models must also become more complex, and therefore difficult and time consuming to create. Even a model of queuing behavior, which is merely one factor in a transaction business, is complex to generate.
SUMMARY OF THE INVENTION
It is the object of the invention to provide an improved method of creating a model of consumer behavior in a complex environment.
According to the invention a method of generating a model of customer behavior in a transaction environment, characterized by the steps of:
selecting a software development tool incorporating at least one artificial life algorithm and capable of constructing a plurality of agents each having at least one drive;
defining at least one drive for each agent which is to be matched to a transaction-related need; and
genetically encoding the defined drives.
Artificial Life or “Alife” was defined by Langton in 1992 as published on http://lslwww.epfl.ch/~moshes/alife.html- “a field of study devoted to understanding life by attempting to abstract the fundamental dynamical principles underlying biological phenomena, and recreating these dynamics in other physical media—such as computers—making them accessible to new kinds of experiment manipulation and testing.
In addition to providing new ways to study the biological phenomena associated with life here on Earth, life-as-we-know-it, Artificial Life allows us to extend our studies to the larger domain of “bio-logic” of possible life, “life-as-it-could-be . . . ”
The drives in Alife systems include hunger, the need to sleep, and the wish to reproduce etc.; within the models, a drive reduction leads to a positive reinforcement in an agent.
In the method according to the invention, at least one such drive is specified and genetically encoded to equate to a consumer need in a transaction environment. Examples include the need for cash, or the need to make a deposit, in a financial transaction environment. Applications in retail or other interactive environments are also possible.
In the method according to the invention, a number of agents are created, each having a plurality of drives and sensors; interfaces are defined between the agents and a representation of a physical environment which is sensed by the sensors, and in which the agents make transactions; for example navigation rules are made which prevent any agent from moving through a wall.
In this specification “agents” means computational systems which inhabit dynamic unpredictable environments.
Also according to the invention, a number of agents is created, for example eight or more, but usually several hundred; the model is run, and, as is known in Alife modelling, a fitness function is applied by which a percentage of the agents is selected which best correspond to observed behavior in real life. In the invention, the agents are selected which best match the observed behavior of humans, for example in a bank branch.
As the model is run, the agents may reproduce or randomly mutate or may remain unchanged; the application of the fitness function may be applied several times until a required number of agents is available which all match actual customer behavior to within predetermined limits.
It is an advantage of a model according to the invention that a full model of each human customer is not required; only the inputs which affect the behavior of a human in a financial transaction, or other selected, environment are required as inputs.
After creation, the model can be constantly updated by comparison with newly derived information about real human customers.
After creation, the model can be used to predict behavior in different environments, such as different bank branch layouts.
REFERENCES:
patent: 5343554 (1994-08-01), Koza et al.
patent: 5708774 (1998-01-01), Boden
patent: 5940816 (1999-08-01), Fuhrer et al.
N. R. Jennings, “Using Intelligent Agents to Manage Business Process”, IEE Colloquium on Intelligent Agents and Their Application, Apr. 1996.*
Han et al, “MultiAgent Problem Solving with Mental States”, IEEE Proceedings of the 1994 Second Australian and New Zealand Conf. on Intelligent Inforation System, Nov.-Dec. 1994.*
Mikami et al, “Combining Reinforcement Learning with GA to find Co-Ordinated Control Rules for Multi-Agent System”, Preceedings of IEEE Inter. Conf. on Evolutionary Computation, May 1996.*
Bimbo et al, “Specification by-Example of Virtual Agents Behavior”, IEEE Transactions on Visualization and Computer Graphics, vol. 1, No. 4, Dec. 1995.*
Thimbleby et al, “Concept of Cooperation in Artificial Life” IEEE Transactions on System, Man and Cybernetics, Jul. 1995.*
Heudin et al, “Artificial Life and Evolutionary Computing in Machine Perception”, IEEE Proceedings of Computer Architectures for Machine Perception, Sep. 1995.*
Wildberger, A. Martin, “Introduction & Overview of “Artificial Life” Evolving Intelligent Agents for Modeling and Simulation”, IEEE Proceedings of the 1996 Winter Simulation Conf., 1996.*
Fulkerson et al, “The Living Factory: Application of Artificial Life to Manufacturing”, IEEE Proceedings of the Second International Symposium on Autonomous Decentralized Systems, Apr. 1995.*
Dave Cliff, “Where Creatures Came From: A review of research literature relevant to CyberLife™ technology, prepared for NCR Financial Systems Ltd.”, School of Cognitive and Computing Sciences, University of Sussex, Brighton BN1 9QH, U.K., Apr. 1997, pp 1-46 (including Bibliography).
Dove Lee G.
Johnson Graham I.
Davis George B.
NCR Corporation
Welte Gregory A.
LandOfFree
Consumer model does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Consumer model, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Consumer model will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3199998