Koninklijke Philips Corporation N.v.

Philips Electronics North America Corporation

Keith E Mathias

J David Schaffer

Larry James Eshelman

Karen Irene Trovato

Murali Mani

System actions such as presentation of information to a user via a remote control device are carried out in accordance with user preferences based on, e.g., observed user behavior over time. The preferences may be learned by observation of user actions and corresponding system state information, and the preferences may be verified by the user. Particular actions may then be repeated under the control of the system, when appropriate conditions are met. Repetition may be automatic, semi-automatic or user initiated. In an illustrative embodiment, information is presented to a user via a remote control device based at least in part on the determined preference. The information presented to the user via the remote control device may include, e.g., an alarm signal associated with a particular parameter specified by the user, a visible icon presented to the user on a display of the remote control device, or other types of information presented via other types of interfaces, such as audio, speech or tactile interfaces.

Method and apparatus for presentation of intelligent, adaptive alarms, icons and other information

The noise associated with conventional techniques for evolutionary improvement of neural network architectures is reduced so that of an optimum architecture can be determined more efficiently and more effectively. Parameters that affect the initialization of a neural network architecture are included within the encoding that is used by an evolutionary algorithm to optimize the neural network architecture. The example initialization parameters include an encoding that determines the initial nodal weights used in each architecture at the commencement of the training cycle. By including the initialization parameters within the encoding used by the evolutionary algorithm, the initialization parameters that have a positive effect on the performance of the resultant evolved network architecture are propagated and potentially improved from generation to generation. Conversely, initialization parameters that, for example, cause the resultant evolved network to be poorly trained, will not be propagated. In accordance with a second aspect of this invention, the encoding also includes parameters that affect the training process, such as the duration of the training cycle, the training inputs applied, and so on. In accordance with a third aspect of this invention, the same set of training or evaluation inputs are applied to all members whose performances are directly compared.

Method for improving neural network architectures using evolutionary algorithms

A security system incorporating a reasoning system and security rules and processes. Transponders may be triggered and sensed from a distance to identify both items and individuals. These sensed identifiers are processed by the reasoning system to determine whether each identified item is authorized to be removed from or brought into a secured location by the identified individual. The system modifies and optimizes its rules and processes based on assessments of security events. The security system enforces these security rules and receives feedback from authorized security personnel. A learning system is configured to modify existing rules or create new rules in conformance with the feedback from the authorized security personnel.

Object proximity/security adaptive event detection

An evolutionary algorithm evolves alternative architectures and parameters for an image classification system. In a preferred embodiment, a learning system is employed, and during the training period of the learning system, the architecture of the learning system is evolved so as to create a learning system that is well suited to the particular classification problem set. In like manner, other parameters of the image classification system are evolved by the evolutionary algorithm, including those that effect image characterization, learning, and classification. An initial set of parameters and architectures are used to create a set of trial classification systems. A number of pre-classified evaluation images are then applied to each system, and each system's resultant classifications for each test case is compared to the proper classification of each test case. Subsequent trial classification systems are evolved based upon the parameters and architecture of the better performing classification systems. The best performing classification system is then selected as the production classification system for classifying new images.

Image classification using evolved parameters

A security system incorporates a reasoning system and security rules and processes that are designed to be as unobtrusive as the situation permits. Two independent aspects of the system facilitate the enforcement of rules and processes in an unobtrusive manner. First, transponders that can be triggered and sensed from a distance are preferably used to identify both items and individuals. These remotely sensed identifiers are processed by the reasoning system to determine whether each identified item is authorized, or likely to be authorized, to be removed from, or brought into, a secured location by the identified individual. Second, the system continually modifies and optimizes its rules and processes based on assessments of security events. An initial set of rules is created for the security system that, generally, prohibit the removal of secured items from the secured location, except that certain individuals are authorized to remove specified items from the secured location. Thereafter, the security system is configured to enforce these security rules and processes, and to receive feedback from authorized security personnel regarding the efficacy of the enforced security rules and processes. Coupled to the security system is a learning system that is configured to modify existing rules or create new rules, in conformance with the feedback from the authorized security personnel. By dynamically adjusting the security rules and processes, the intrusion of the security system on the monitored individuals is substantially reduced, and the system continues to be optimized based on continued feedback.

A code compaction based on macro substitutions is presented wherein the choice of possible macro substitutions is guided by an evolutionary algorithm process. In a preferred embodiment, a random population of sets of macro substitutions are generated and a compaction effectiveness is evaluated for each set. This random population is partitioned into pairs of &ldquo;parents&rdquo;, and each pair of parents produce a pair of &ldquo;offspring&rdquo;. The effectiveness of the compaction provided by each of the offspring is evaluated, and a &ldquo;survival of the fittest&rdquo; algorithm is applied to identify the individuals that have the best compaction effectiveness. These preferred individuals are partitioned into pairs of parents who produce pairs of offspring, and the most compaction-effective individuals are selected to be parents for the next generation. This process continues until subsequent generations show insignificant improvement, and the best individual is selected as the solution.

Code compaction by evolutionary algorithm

A user interface is provided that allows for the control of the movement of multiple objects by the identification of a target location associated with each of the multiple objects. The identification of the target location is via a coordinate pointer associated with an input device, and the association of the target location to a particular object is via the closure of a switch on an input device that is associated with the particular object. After identifying an associated target location, the object is moved toward the target location without further input from the user. In a preferred embodiment, the movement of the object toward the target location is dependent upon rules of motion and rules of engagement with other objects or obstacles, thereby facilitating a realistic rendering of the object's path toward the target. The rules of motion and engagement may be specific to each type of object being moved, and also subject to the interactions of multiple objects and obstacles.

Growing community of inventors

Ossining, NY, United States of America

Keith E Mathias