Patents Assigned to PROGRESS, INC.
-
Publication number: 20200019862Abstract: A neural network includes inputs for receiving input signals, synapses connected to the inputs and having corrective weights, and neurons having outputs connected with the inputs via the synapses. Each neuron generates a neuron sum by summing corrective weights selected from the respective synapse. A controller receives a desired output signal, determines a deviation of the neuron sum from the desired output signal value, and modifies respective corrective weights using the determined deviation. Adding up the modified corrective weights to determine the neuron sum minimizes the deviation and trains the network. A structure-forming module rearranges connections between network elements during the training and a signal allocation module distributes the input signals among the network elements during the training.Type: ApplicationFiled: July 26, 2019Publication date: January 16, 2020Applicant: Progress, Inc.Inventors: Dmitri Pescianschi, Vladimir Proseanic, Boris Zlotin
-
Patent number: 10423694Abstract: A neural network includes inputs for receiving input signals, and synapses connected to the inputs and having corrective weights organized in an array. Training images are either received by the inputs as an array or codified as such during training of the network. The network also includes neurons, each having an output connected with at least one input via one synapse and generating a neuron sum array by summing corrective weights selected from each synapse connected to the respective neuron. Furthermore, the network includes a controller that receives desired images in an array, determines a deviation of the neuron sum array from the desired output value array, and generates a deviation array. The controller modifies the corrective weight array using the deviation array. Adding up the modified corrective weights to determine the neuron sum array reduces the subject deviation and generates a trained corrective weight array for concurrent network training.Type: GrantFiled: March 3, 2017Date of Patent: September 24, 2019Assignee: Progress, Inc.Inventor: Dmitri Pescianschi
-
Publication number: 20170177998Abstract: A neural network includes inputs for receiving input signals, and synapses connected to the inputs and having corrective weights organized in an array. Training images are either received by the inputs as an array or codified as such during training of the network. The network also includes neurons, each having an output connected with at least one input via one synapse and generating a neuron sum array by summing corrective weights selected from each synapse connected to the respective neuron. Furthermore, the network includes a controller that receives desired images in an array, determines a deviation of the neuron sum array from the desired output value array, and generates a deviation array. The controller modifies the corrective weight array using the deviation array. Adding up the modified corrective weights to determine the neuron sum array reduces the subject deviation and generates a trained corrective weight array for concurrent network training.Type: ApplicationFiled: March 3, 2017Publication date: June 22, 2017Applicant: Progress, Inc.Inventor: Dmitri Pescianschi
-
Patent number: 9619749Abstract: A neural network includes a plurality of inputs for receiving input signals, and synapses connected to the inputs and having corrective weights established by a memory element that retains a respective weight value. The network additionally includes distributors. Each distributor is connected to one of the inputs for receiving the respective input signal and selects one or more corrective weights in correlation with the input value. The network also includes neurons. Each neuron has an output connected with at least one of the inputs via one synapse and generates a neuron sum by summing corrective weights selected from each synapse connected to the respective neuron. The output of each neuron provides the respective neuron sum to establish operational output signal of the network. A method of operating a neural network includes processing data thereby and using modified corrective weight values established by a separate analogous neural network during training thereof.Type: GrantFiled: June 9, 2016Date of Patent: April 11, 2017Assignee: Progress, Inc.Inventor: Dmitri Pescianschi
-
Publication number: 20160283842Abstract: A neural network includes a plurality of inputs for receiving input signals, and synapses connected to the inputs and having corrective weights established by a memory element that retains a respective weight value. The network additionally includes distributors. Each distributor is connected to one of the inputs for receiving the respective input signal and selects one or more corrective weights in correlation with the input value. The network also includes neurons. Each neuron has an output connected with at least one of the inputs via one synapse and generates a neuron sum by summing corrective weights selected from each synapse connected to the respective neuron. The output of each neuron provides the respective neuron sum to establish operational output signal of the network. A method of operating a neural network includes processing data thereby and using modified corrective weight values established by a separate analogous neural network during training thereof.Type: ApplicationFiled: June 9, 2016Publication date: September 29, 2016Applicant: Progress, Inc.Inventor: Dmitri Pescianschi
-
Patent number: 9390373Abstract: A neural network includes a plurality of inputs for receiving input signals, and synapses connected to the inputs and having corrective weights. The network additionally includes distributors. Each distributor is connected to one of the inputs for receiving the respective input signal and selects one or more corrective weights in correlation with the input value. The network also includes neurons. Each neuron has an output connected with at least one of the inputs via one synapse and generates a neuron sum by summing corrective weights selected from each synapse connected to the respective neuron. Furthermore, the network includes a weight correction calculator that receives a desired output signal, determines a deviation of the neuron sum from the desired output signal value, and modifies respective corrective weights using the determined deviation. Adding up the modified corrective weights to determine the neuron sum minimizes the subject deviation for training the neural network.Type: GrantFiled: September 23, 2015Date of Patent: July 12, 2016Assignee: Progress, Inc.Inventor: Dmitri Pescianschi
-
Publication number: 20160012330Abstract: A neural network includes a plurality of inputs for receiving input signals, and synapses connected to the inputs and having corrective weights. The network additionally includes distributors. Each distributor is connected to one of the inputs for receiving the respective input signal and selects one or more corrective weights in correlation with the input value. The network also includes neurons. Each neuron has an output connected with at least one of the inputs via one synapse and generates a neuron sum by summing corrective weights selected from each synapse connected to the respective neuron. Furthermore, the network includes a weight correction calculator that receives a desired output signal, determines a deviation of the neuron sum from the desired output signal value, and modifies respective corrective weights using the determined deviation. Adding up the modified corrective weights to determine the neuron sum minimizes the subject deviation for training the neural network.Type: ApplicationFiled: September 23, 2015Publication date: January 14, 2016Applicant: PROGRESS, INC.Inventor: Dmitri Pescianschi
-
Patent number: 8954456Abstract: A system and method for accepting content in a first form, identifying the type of the content, specifying at least a second form, converting the content form from the first to the second form, and continuing conversion for any forms subsequent to the second form. Variant Production Functions create specification-based variant files and supplemental metadata using data from existing resources. A Variant Production Function provides functionality for parsing an existing resource's data, and generating a complete or partial alternate translated form of that resource. The Variant Production Function includes at least the input of the content of the primary target resource file used as the basis for the new variant. Variant Production Functions may be in Auto Generation Configuration Objects, or used as components within a Translation and Transcription Service.Type: GrantFiled: March 26, 2014Date of Patent: February 10, 2015Assignee: Measured Progress, Inc.Inventors: Zachary Pierce, Thomas Hoffmann, Michael K. Russell
-
Patent number: 8303309Abstract: In an exemplary embodiment, a computer-implemented method for implementing a test delivery system is provided. The method includes implementing a plurality of interoperable accommodation and accessibility tools, and providing an interface to the test with the plurality of interoperable accommodation and accessibility tools provided on the interface. The method also includes receiving one or more tool settings for a test-taker specifying one or more permitted tools in the plurality of interoperable accommodation and accessibility tools, and customizing the interface based on the one or more tool settings to allow the test-taker access to the one or more permitted tools during the test. The method further includes allowing the test-taker to activate or deactivate each of the one or more permitted tools.Type: GrantFiled: July 11, 2008Date of Patent: November 6, 2012Assignee: Measured Progress, Inc.Inventors: Thomas Hoffmann, Michael K. Russell
-
Patent number: 7202781Abstract: There is provided a monitoring system for remotely monitoring an item or activity, the monitoring system including a recording device for recording a visual record of the item, a communicator for communicating the visual record, and a security device for controlling access to the monitoring system. Also provided is a method of remotely monitoring an item by recording a digital image of the item, communicating the digital image of the item through a communication system, and accessing the digital image using a secured access device.Type: GrantFiled: October 29, 2003Date of Patent: April 10, 2007Assignee: See Progress, Inc.Inventors: David Henderson, Gordon Henderson, Jerry Henderson
-
Patent number: 7151448Abstract: There is provided a monitoring system for remotely monitoring an item or activity, the monitoring system including a recording device for recording a visual record of the item, a communicator for communicating the visual record, and a security device for controlling access to the monitoring system. Also provided is a method of remotely monitoring an item by recording a digital image of the item, communicating the digital image of the item through a communication system, and accessing the digital image using a secured access device.Type: GrantFiled: October 16, 2002Date of Patent: December 19, 2006Assignee: See Progress, Inc.Inventors: David Henderson, Gordon Henderson, Jerry Henderson
-
Publication number: 20050086257Abstract: A system and process includes: a relational database that has a core database of global rules and project data, an item tracking database, a scanning and image management database, and a scoring database; an enterprise integrated application software program and computer network; and a web interface linking host operations and users; all of which is used for generating a production order and packing schedule of test documents from the rules and project data, for tracking production, packing, shipping, return shipping, and processing of actual test documents by using unique identifiers assigned to every document and monitoring and recording every movement of every document and appending all records of movement and other changes of condition to the original project data in the relational database, for scanning all documents into image form and storing and scoring the images, and for generating reports based on the scores and demographics of the testees.Type: ApplicationFiled: October 15, 2004Publication date: April 21, 2005Applicant: MEASURED PROGRESS, INC.Inventor: Lonnie Wright
-
Publication number: 20040229199Abstract: A computer-based testing system is disclosed comprising: a data administration system including centrally hosted data administration servers; a network and an operational testing system the data administration system including a browser-capable workstation connectible via the network to the centrally hosted data administration servers. The operational testing system may include three subsystems connected to the network: a test delivery server running on a test delivery workstation and managing all aspects of a test session by acting as a data repository and hub for communication between the other subsystems, a proctor software running on a proctor test workstation providing a user interface for managing a test session by communicating with the test delivery server, and a student test software running on a student test workstation providing a user interface for displaying test items and recording responses.Type: ApplicationFiled: April 15, 2004Publication date: November 18, 2004Applicant: MEASURED PROGRESS, INC.Inventors: Edmund P. Ashley, R. Craig Enslin, Neal M. Kingston, Chloe Y. Torres, David G. Wozmak, Michael G. Willett
-
Patent number: 6511326Abstract: An evaluation method presents a series of question to an evaluated person. The questions represent a matrix of tracks and items. The questions are presented in order of increasing difficulty along a selected track until a first condition occurs. After the first condition occurs, a new track is selected, and the questions along the newly selected track are presented in order of increasing difficulty until a second condition occurs. The new track selection and question presentation are repeated until a third condition occurs. The questions may be defined by a prime question and a series of subcomponents, which represent alternate presentations of the prime question. The results of the evaluation performed are assembled from sentence fragments corresponding to the prime question and series of subcomponents and are presented in a narrative form.Type: GrantFiled: June 27, 2000Date of Patent: January 28, 2003Assignee: Children's Progress, Inc.Inventors: Eugene Galanter, Michelle Galanter
-
Patent number: 5407464Abstract: The present invention relates to ultrafine comminution of mineral and organic powders with the aid of ball-milling techniques employing, as the grinding medium, solid (nonporous) metal-carbide microspheres with diameters of from 10 to 100 .mu.m fabricated with the aid of high-temperature plasma-torch reactors from fully liquefied carbides of tungsten, thallium, niobium and vanadium.Type: GrantFiled: January 12, 1994Date of Patent: April 18, 1995Assignee: Industrial Progress, Inc.Inventor: Adam F. Kaliski
-
Patent number: 5378399Abstract: Complex (multicomponent) functional microgels with rapid formation kinetics and compositions thereof, as well as process for synthesizing said complex microgels in aqueous media.Type: GrantFiled: July 27, 1992Date of Patent: January 3, 1995Assignee: Industrial Progress, Inc.Inventor: Adam F. Kaliski
-
Patent number: 5346546Abstract: Aggregate-TiO.sub.2 pigment products comprising in each 100 parts, by weight, at least 50 parts, by weight, particulate TiO.sub.2 materials bound intrinsically with other pigmentary, subpigmentary and nonpigmentary components with the aid of inorganic and/or organic cements/adhesives.Type: GrantFiled: March 23, 1992Date of Patent: September 13, 1994Assignee: Industrial Progress, Inc.Inventor: Adam F. Kaliski
-
Patent number: 5312484Abstract: Composite pigment products containing from 4.5% up to 50%, by weight, of particulate titanium dioxide in which particles of the latter and other particulate pigmentary, subpigmentary and nonpigmentary components are cemented intrinsically with the aid of in-situ synthesized complex microgels.Type: GrantFiled: December 23, 1991Date of Patent: May 17, 1994Assignee: Industrial Progress, Inc.Inventor: Adam F. Kaliski
-
Patent number: 5299674Abstract: A mobile belt support car for a conveyor system having an extensible conveyor belt which includes a troughing strand and a return strand. Each car of a mobile support structure includes a longitudinal frame having first and second opposite axial ends, first and second horizontally pivoting joints provided respectively at the first and second opposite axial ends of the longitudinal frame, a transverse beam fixedly connected to the longitudinal frame and having first and second opposite axial ends, first and second wheels supported respectively at the first and second opposite axial ends of the transverse beam, and a first idler frame assembly mounted on the longitudinal frame and including an upper cantilevered beam, a lower cantilevered beam, the first and second cantilevered beams extending towards the extensible conveyor belt, and upper rollers, mounted on the upper cantilevered beam, for supporting the troughing strand of the extensible conveyor belt.Type: GrantFiled: December 20, 1991Date of Patent: April 5, 1994Assignee: Westfalia Mining Progress, Inc.Inventor: Wayne S. Cusitar
-
Patent number: 5240561Abstract: A fundamentally new papermaking process for the manufacture of paper, board and other wet-laid products on a paper machine under conditions ranging from acidic to alkaline from aqueous furnishes treated with in-situ-synthesized complex functional microgels.Type: GrantFiled: August 20, 1992Date of Patent: August 31, 1993Assignee: Industrial Progress, Inc.Inventor: Adam E. Kaliski