Abstract: A method of using a computing device to manage a lifecycle of machine learning models includes receiving, by a computing device, multiple pre-defined machine learning lifecycle tasks. The computing device manages executing a management-layer software layer for the multiple pre-defined machine learning lifecycle tasks. The computing device further generates and updates a machine learning pipeline using the management-layer software layer.

Abstract: Noisy labeled and unlabeled datapoint detection and rectification in a training dataset for machine-learning is facilitated by a processor(s) obtaining a training dataset for use in training a machine-learning model. The processor(s) applies ensemble machine-learning and a generative model to the training dataset to detect noisy labeled datapoints in the training dataset, and create a clean dataset with preliminary labels added for any unlabeled datapoints in the training dataset. Data-driven active learning and the clean dataset are used by the processor(s) to facilitate generating an active-learned dataset with true labels added for one or more selected datapoints of a datapoint pool including the detected noisy labeled datapoints and the unlabeled datapoints of the training dataset. The machine-learning model is trained by the processor(s) using, at least in part, the clean dataset and the active-learned dataset.

Abstract: A method of using a computing device to manage a lifecycle of machine learning models includes receiving, by a computing device, multiple pre-defined machine learning lifecycle tasks. The computing device manages executing a management-layer software layer for the multiple pre-defined machine learning lifecycle tasks. The computing device further generates and updates a machine learning pipeline using the management-layer software layer.

Abstract: Noisy labeled and unlabeled datapoint detection and rectification in a training dataset for machine-learning is facilitated by a processor(s) obtaining a training dataset for use in training a machine-learning model. The processor(s) applies ensemble machine-learning and a generative model to the training dataset to detect noisy labeled datapoints in the training dataset, and create a clean dataset with preliminary labels added for any unlabeled datapoints in the training dataset. Data-driven active learning and the clean dataset are used by the processor(s) to facilitate generating an active-learned dataset with true labels added for one or more selected datapoints of a datapoint pool including the detected noisy labeled datapoints and the unlabeled datapoints of the training dataset. The machine-learning model is trained by the processor(s) using, at least in part, the clean dataset and the active-learned dataset.

Abstract: Crosstalk in a piezo printhead is reduced by selecting an actuation signal for a nozzle, determining a time delay and a pulse width extension based on adjacent actuation signals of adjacent nozzles, and applying the time delay and pulse width extension to the actuation signal.

Abstract: A method for updating an application on a host system includes steps or acts of: installing an application on the host system; installing a virtual machine monitor on the host system, installing a first virtual container on the host system, wherein the first virtual container comprises at least one update to the installed application; and instantiating the first virtual container under the virtual machine monitor in a mode wherein the host system can be accessed but not modified and wherein instantiating the first virtual container includes updating the installed application in the first virtual computer and running the updated application in the first virtual container. The method may also include a step of confirming that the updated application runs properly. Optionally, the virtual container may be devirtualized.

Abstract: Crosstalk in a piezo printhead is reduced by selecting an actuation signal for a nozzle, determining a time delay and a pulse width extension based on adjacent actuation signals of adjacent nozzles, and applying the time delay and pulse width extension to the actuation signal.

Abstract: Disclosed is a method, computer program product and a server unit of a virtual execution environment having at least one client. The server unit includes a network interface for conducting bidirectional communications with the at least one client and a control unit configured to determine, for individual ones of positions in a virtualized execution of an application, those components that are required at each position; to establish a pseudo-distance between the positions in the virtualized execution of the application; and to identify components that are candidates for prefetching as those components required at one or more positions that are proximate to a current position. The pseudo-distance between positions is indicative of an urgency of prefetching candidate components for regulating an amount of resources devoted to the prefetching of components.

Abstract: Disclosed is a method, apparatus and a computer program of a virtual execution environment. In one aspect thereof a data processor includes a read-only storage medium and a virtual machine monitor configured to execute a program in a virtual container. Prior to execution of the application a set of application components accessed during a phase change are recorded in the read-only storage medium in a manner predetermined to minimize the time needed to retrieve the set of components. At the occurrence of the phase change, or prior to the occurrence of a phase change if the files are prefetched, the virtual machine monitor retrieves the components from the read-only storage medium. In a further aspect thereof there is provided a packager of an application to be executed in a virtualization environment by a virtual machine monitor.

Abstract: Disclosed is a method, apparatus and a computer program of a virtual execution environment. In one aspect thereof a data processor includes a disk-based storage system and a virtual machine monitor configured to execute a program in a virtual container. The virtual machine monitor is responsive, prior to execution of the application, to record a set of files accessed during a phase change in the disk-based storage system in a manner predetermined to minimize the time needed to retrieve the set of files. The virtual machine monitor is further responsive, at the occurrence of the phase change, or prior to the occurrence of a phase change if the files are prefetched, to retrieve the files from the disk-based storage system.

Abstract: An apparatus and method for steering a beam of light using an array of tunable optical phase delay elements is presented. The sidelobes of an angular spectrum of light reflected from the array are causing an optical crosstalk. The selected sidelobes are suppressed by perturbing the phase delay pattern of the array elements. The pattern of perturbations is found by linearizing a system of equations describing dependence of the angular spectrum of the reflected light on the phase delays introduced into the wavefront of light by the elements of the array.

Abstract: The present invention provides a method, apparatus, and computer instructions for program analysis. In an exemplary embodiment, a target program is loaded onto a data processing system including a program analysis software component. Hierarchies in the form of control flow and containment information about the target program are determined, using either static or dynamic, and direct (e.g., instrumented profilers) or indirect (e.g., monitored runtime), processing techniques, depending on the application and program analysis tool being used. The control flow and object containment information are then presented in a related and selected format for consideration by a user. One embodiment includes a program and GUI (graphical user interface) operably providing control flow-annotated object containment hierarchies to a user, the hierarchies including control flow, allocation and storage information added to object containment information.

Abstract: An apparatus and method for steering a beam of light using an array of tunable optical phase delay elements is presented. The sidelobes of an angular spectrum of light reflected from the array are causing an optical crosstalk. The selected sidelobes are suppressed by perturbing the phase delay pattern of the array elements. The pattern of perturbations is found by linearizing a system of equations describing dependence of the angular spectrum of the reflected light on the phase delays introduced into the wavefront of light by the elements of the array.

Abstract: Disclosed is a method, apparatus and a computer program of a virtual execution environment. In one aspect thereof a data processor includes a disk-based storage system and a virtual machine monitor configured to execute a program in a virtual container. The virtual machine monitor is responsive, prior to execution of the application, to record a set of files accessed during a phase change in the disk-based storage system in a manner predetermined to minimize the time needed to retrieve the set of files. The virtual machine monitor is further responsive, at the occurrence of the phase change, or prior to the occurrence of a phase change if the files are prefetched, to retrieve the files from the disk-based storage system.

Abstract: Disclosed is a method, apparatus and a computer program of a virtual execution environment. In one aspect thereof a data processor includes a read-only storage medium and a virtual machine monitor configured to execute a program in a virtual container. Prior to execution of the application a set of application components accessed during a phase change are recorded in the read-only storage medium in a manner predetermined to minimize the time needed to retrieve the set of components. At the occurrence of the phase change, or prior to the occurrence of a phase change if the files are prefetched, the virtual machine monitor retrieves the components from the read-only storage medium. In a further aspect thereof there is provided a packager of an application to be executed in a virtualization environment by a virtual machine monitor.

Abstract: Disclosed is a method, computer program product and a server unit of a virtual execution environment having at least one client. The server unit includes a network interface for conducting bidirectional communications with the at least one client and a control unit configured to determine, for individual ones of positions in a virtualized execution of an application, those components that are required at each position; to establish a pseudo-distance between the positions in the virtualized execution of the application; and to identify components that are candidates for prefetching as those components required at one or more positions that are proximate to a current position. The pseudo-distance between positions is indicative of an urgency of prefetching candidate components for regulating an amount of resources devoted to the prefetching of components.

Abstract: A method for updating an application on a host system includes steps or acts of: installing an application on the host system; installing a virtual machine monitor on the host system, installing a first virtual container on the host system, wherein the first virtual container comprises at least one update to the installed application; and instantiating the first virtual container under the virtual machine monitor in a mode wherein the host system can be accessed but not modified and wherein instantiating the first virtual container includes updating the installed application in the first virtual computer and running the updated application in the first virtual container. The method may also include a step of confirming that the updated application runs properly. Optionally, the virtual container may be devirtualized.

Abstract: A method for setting configuration parameters for at least one software system, comprises the steps of: a) receiving an identification of a set of configuration parameters for at least one software system to be optimized; b) selecting a random value from a predetermined range for each configuration parameter of interest; c) setting each configuration parameter to a corresponding random value selected; d) running an application using the values selected; e) gathering performance information relating to the software system while the application is running; f) repeating steps b) through e) for a selected number of times; and g) performing an analysis of the performance information gathered to determine optimal configuration parameters. The method can be performed by a programmable computer system running program instructions for carrying out the above method steps or by a specialized apparatus such as an ASIC.

Abstract: A method for setting configuration parameters for at least one software system, comprises the steps of: a) receiving an identification of a set of configuration parameters for at least one software system to be optimized; b) selecting a random value from a predetermined range for each configuration parameter of interest; c) setting each configuration parameter to a corresponding random value selected; d) running an application using the values selected; e) gathering performance information relating to the software system while the application is running; f) repeating steps b) through e) for a selected number of times; and g) performing an analysis of the performance information gathered to determine optimal configuration parameters. The method can be performed by a programmable computer system running program instructions for carrying out the above method steps or by a specialized apparatus such as an ASIC.