Abstract:
A system is provided for benchmarking a progressive combinatorial solver. The system may initialize a parametric model collector, and perform a plurality of parametric trials associated with respective numbers of sub-part iterations of the progressive combinatorial solver. For each of the plurality of parametric trials, the system may initialize a statistical data collector. The system may perform a plurality of randomized executions of the progressive combinatorial solver, and add data including metric values for respective solutions of the executions to the statistical data collector. The system may then determine statistical parameter(s) of data in the statistical data collector, and add data including an indication of the respective number of sub-part iterations, a cost of finding the metric values for the respective solutions, and the statistical parameters of the trial to the parametric data collector. And the system may determine a best-fit model from data in the parametric data collector.
Abstract:
Systems, methods, and apparatus for a multi-spectral X-ray target and source are disclosed. In one or more embodiments, a disclosed method comprises emitting, by a source of the X-ray generator, electrons towards a section of a multi-spectral X-ray target of the X-ray generator. In one or more embodiments, the multi-spectral X-ray target is rotatable and comprises a plurality of sections, which each comprise an X-ray generating material and at least two of the sections comprise a different X-ray generating material. The method further comprises generating a set of X-rays, when the electrons impinge on the section of the multi-spectral X-ray target. The method further comprises rotating the multi-spectral X-ray target such that the source is in position to project the electrons towards another section of the multi-spectral X-ray target. Further, the method comprises repeating the above method steps for all of the remaining sections of the multi-spectral X-ray target.
Abstract:
A system is provided for benchmarking a progressive combinatorial solver. The system may initialize a parametric model collector, and perform a plurality of parametric trials associated with respective numbers of sub-part iterations of the progressive combinatorial solver. For each of the plurality of parametric trials, the system may initialize a statistical data collector. The system may perform a plurality of randomized executions of the progressive combinatorial solver, and add data including metric values for respective solutions of the executions to the statistical data collector. The system may then determine statistical parameter(s) of data in the statistical data collector, and add data including an indication of the respective number of sub-part iterations, a cost of finding the metric values for the respective solutions, and the statistical parameters of the trial to the parametric data collector. And the system may determine a best-fit model from data in the parametric data collector.
Abstract:
A system is provided for benchmarking a progressive combinatorial solver. The system may initialize a parametric model collector, and perform a plurality of parametric trials associated with respective numbers of sub-part iterations of the progressive combinatorial solver. For each of the plurality of parametric trials, the system may initialize a statistical data collector. The system may perform a plurality of randomized executions of the progressive combinatorial solver, and add data including metric values for respective solutions of the executions to the statistical data collector. The system may then determine statistical parameter(s) of data in the statistical data collector, and add data including an indication of the respective number of sub-part iterations, a cost of finding the metric values for the respective solutions, and the statistical parameters of the trial to the parametric data collector. And the system may determine a best-fit model from data in the parametric data collector.
Abstract:
Systems, methods, and apparatus for a multi-spectral X-ray target and source are disclosed. In one or more embodiments, a disclosed method comprises emitting, by a source of the X-ray generator, electrons towards a section of a multi-spectral X-ray target of the X-ray generator. In one or more embodiments, the multi-spectral X-ray target is rotatable and comprises a plurality of sections, which each comprise an X-ray generating material and at least two of the sections comprise a different X-ray generating material. The method further comprises generating a set of X-rays, when the electrons impinge on the section of the multi-spectral X-ray target. The method further comprises rotating the multi-spectral X-ray target such that the source is in position to project the electrons towards another section of the multi-spectral X-ray target. Further, the method comprises repeating the above method steps for all of the remaining sections of the multi-spectral X-ray target.
Abstract:
A system is provided for benchmarking a progressive combinatorial solver. The system may initialize a parametric model collector, and perform a plurality of parametric trials associated with respective numbers of sub-part iterations of the progressive combinatorial solver. For each of the plurality of parametric trials, the system may initialize a statistical data collector. The system may perform a plurality of randomized executions of the progressive combinatorial solver, and add data including metric values for respective solutions of the executions to the statistical data collector. The system may then determine statistical parameter(s) of data in the statistical data collector, and add data including an indication of the respective number of sub-part iterations, a cost of finding the metric values for the respective solutions, and the statistical parameters of the trial to the parametric data collector. And the system may determine a best-fit model from data in the parametric data collector.