Distributed evolutionary algorithm for asset management and trading

• US 8,825,560 B2
• Filed: 05/15/2013
• Issued: 09/02/2014
• Est. Priority Date: 11/08/2007
• Status: Active Grant

• Current Assignee: Cognizant Technology Solutions U.S. Corporation (Cognizant Technology Solutions Corp.)
• Original Assignee: Genetic Finance (Barbados) Limited
• Inventors: Hodjat, Babak, Shahrzad, Hormoz, Blondeau, Antoine, Cheyer, Adam, Harrigan, Peter
• Primary Examiner(s): CHEN, ALAN S
• Application Number:
  US13/895,238
• Publication Number:
  US 20130254142A1
• Time in Patent Office: 475 Days
• Field of Search: 706/10, 706/12
• US Class Current: 706/10
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/086 using evolutionary algorithms, e.g. genetic algorithms or genetic programming
• CPC Class Codes: G06N 3/12 using genetic models
• CPC Class Codes: G06N 3/126 Evolutionary algorithms, e.g. genetic algorithms or genetic programming
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A server computer and a multitude of client computers form a network computing system that is scalable and adapted to continue to evaluate the performance characteristics of a number of genes generated using a software application running on the client computers. Each client computer continues to periodically receive data associated with the genes stored in its memory. Using this data, the client computers evaluate the performance characteristic of their genes by comparing a solution provided by the gene with the periodically received data associated with that gene. Accordingly, the performance characteristic of each gene may be updated and varied with each periodically received data. The performance characteristic of a gene defines its fitness. The genes may be virtual asset traders that recommend trading options, and the data associated with the genes may be historical trading data.

Claims

1. A computer-implemented data mining system, comprising:

a data processor; and

a memory accessible to the data processor and identifying a candidate database having a pool of individuals, each of the individuals further having associated therewith an indication of a respective fitness estimate,

for use with a data mining training database accessible to the data processor and identifying training data,

the data processor configured to;

test on a first subset of the training data each individual in a testing subset of at least one of the individuals;

calculate a fitness estimate for each of the individuals in the testing subset in dependence upon the tests on the first subset of the training data;

discard individuals from the pool of individuals in dependence upon their fitness estimates;

test on a second subset of the training data, each individual in the testing subset which remains after the discarding of individuals, the second subset including at least one datum not included in the first subset;

update, in dependence upon the tests on the second subset of the training data, the fitness estimate for each individual tested on the second subset of the training data; and

discard further individuals from the pool of individuals in dependence upon their updated fitness estimates.

2. A system according to claim 1, wherein the first subset of the training data also includes at least one datum not included in the second subset of the training data.

3. A system according to claim 1, wherein each individual in the pool identifies a respective set of at least one condition and at least one corresponding proposed output in dependence upon the conditions.

4. A system according to claim 3 wherein the training database contains historical market data for one or more financial assets,

wherein each individual indicates a plurality of conditions resolvable in dependence upon the market data in the training database, the outputs identified by the individuals including trading instructions for one or more of the financial assets,

and wherein the data processor is further configured to provide, for use in generating trading recommendations in dependence upon production data, individuals from the pool of individuals not discarded from the pool in dependence upon their updated fitness estimates.

5. A system according to claim 1, wherein the data processor is further configured to transmit to a server computer system individuals remaining after the discarding of further individuals from the pool of individuals in dependence upon their updated fitness estimates.

6. A system according to claim 5, further comprising the server computer system, wherein the server computer system comprises:

a server data processor; and

a memory accessible to the data processor, the server data processor configured to transmit selected received individuals to a client computer system for further testing on the training data.

7. A system according to claim 6, wherein the server data processor is further configured to discard selected received individuals in dependence upon their fitness estimates.

8. A system according to claim 1, wherein the data processor includes first and second computer systems,

wherein the first computer system is configured to perform the testing on the first subset of the training data and the second computer system is configured to perform the discarding of individuals in dependence upon their fitness estimates.

9. A system according to claim 1, wherein the data processor is further configured to produce new individuals by reproduction in dependence upon one or more individuals remaining after the discarding of further individuals.

10. A system according to claim 1, wherein the discarding of individuals from the pool of individuals in dependence upon their fitness estimates, operates further in dependence upon the quantity of testing that has been experienced by the individuals.