Abstract: A method for dynamically assigning an inference request is disclosed. A method for dynamically assigning an inference request may include determining at least one model to process an inference request on a plurality of computing platforms, the plurality of computing platforms including at least one Central Processing Unit (CPU) and at least one Graphics Processing Unit (GPU), obtaining, with at least one processor, profile information of the at least one model, the profile information including measured characteristics of the at least one model, dynamically determining a selected computing platform from between the at least one CPU and the at least one GPU for responding to the inference request based on an optimized objective associated with a status of the computing platform and the profile information, and routing, with at least one processor, the inference request to the selected computing platform. A system and computer program product are also disclosed.

Abstract: Provided are systems for ensemble learning with machine learning models that include a processor to receive a training dataset of a plurality of data instances, wherein each data instance comprises a time series of data points, add an amount of time delay to one or more data instances to provide an augmented training dataset, select a first plurality of supervised machine learning models, select a second plurality of unsupervised machine learning models, train the first plurality of supervised machine learning models and the second plurality of unsupervised machine learning models based on the augmented training dataset, generate an ensemble machine learning model based on outputs of the supervised machine learning models and unsupervised machine learning models, and generate a runtime output of the ensemble machine learning model based on a runtime input to the ensemble machine learning model. Methods and computer program products are also provided.

Abstract: Described are a method, system, and computer program product for operating dynamic shadow testing environments for machine-learning models. The method includes generating a shadow testing environment operating at least two transaction services. The method also includes receiving a plurality of transaction authorization requests. The method further includes determining a first percentage associated with a first testing policy of the first transaction service and a second percentage associated with a second testing policy of the second transaction service. The method further includes replicating in the shadow testing environment, in real-time with processing the payment transactions, a first portion of the plurality of transaction authorization requests and a second portion of the plurality of transaction authorization requests.

Abstract: A hydrogen production power supply system is provided by the present disclosure, wherein two stage conversion modules are arranged between the power supplies, and each output end of the second stage conversion module is connected to a power supply end of each hydrogen production device, respectively, so that independent power supply is provided for each hydrogen production device, which eliminates the loss of gas production caused by the non-uniform load among the hydrogen production devices. In addition, being different from the conventional technology, it is unnecessary for the hydrogen production power supply system to adapt the voltage by using a customized power frequency converter in the present disclosure. Therefore, the path loss issue can be avoided.

Abstract: A method may include obtaining feature data associated with a set of feature inputs of a machine learning model, determining a probability that a subset of the feature data is concurrently used as the set of feature inputs for the machine learning model, and storing the subset of the feature data on a same cache node or server of a plurality of cache nodes or servers based on the probability.

Abstract: A booting method is provided, applied to a computer system. The computer system includes a transmission interface, a power key, and a trigger element. The transmission interface includes a transmission specification and is electrically connected to a graphics card, and the power key is used for driving the computer system to perform a booting procedure. The booting method includes: detecting, in the booting procedure, whether the trigger element is triggered or not; and lowering the transmission specification and restarting the computer system when the trigger element is triggered. A computer system adopting the booting method is further provided.

Abstract: Systems and methods of the present disclosure provide techniques to locate target devices. An electronic message is sent to a set of landmark devices, signaling each landmark device to transmit echo-request packets to a target device and to other landmark devices and measure network-communication delays. Indications of the network-communication delays are received from the set landmark devices. A first sequence is formed by sorting the set of landmark devices relative to the network-communication delays between the target device and each landmark device. For each respective landmark device, an additional sequence is formed by sorting other landmark devices relative to the network-communication delays between the respective landmark device and the other landmark devices. A sequence-matching operation is applied to the first sequence and the additional sequences to form a ranking of the set of landmark devices relative to the target device.

Abstract: A method, system, and computer program product for dynamically scheduling machine learning inference jobs receive or determine a plurality of performance profiles associated with a plurality of system resources, wherein each performance profile is associated with a machine learning model; receive a request for system resources for an inference job associated with the machine learning model; determine a system resource of the plurality of system resources for processing the inference job associated with the machine learning model based on the plurality of performance profiles and a quality of service requirement associated with the inference job; assign the system resource to the inference job for processing the inference job; receive result data associated with processing of the inference job with the system resource; and update based on the result data, a performance profile of the plurality of the performance profiles associated with the system resource and the machine learning model.

Abstract: Described are a system, method, and computer program product for operating dynamic shadow testing environments for machine-learning models. The method includes storing a testing policy including an identifier of a machine-learning model and an identifier of a transaction service. The method includes generating a shadow testing environment operating the transaction service using the machine-learning model. The method also includes receiving, at a transaction service provider system, a transaction authorization request including transaction data of a transaction associated with a payment device. The method further includes identifying the machine-learning model associated with the transaction based on a parameter of the transaction data. The method further includes determining, based on the identifier of the machine-learning model, the testing policy and the shadow testing environment.

Abstract: A multi-channel alkaline hydrogen production system is disclosed. Using liquid outlets of a hydrogen alkali treatment unit and an oxygen alkali treatment unit, a circulating alkaline liquid is outputted to an alkaline liquid circulating pump and a controllable channel, and then the circulating alkaline liquid is returned to the negative electrode of an electrolyzer; Thus, a controller can control the amount of produced hydrogen according to the measured current of the electrolyzer, then calculates a corresponding alkaline liquid circulation volume reference value according to the amount of produced hydrogen, and according to the alkaline liquid circulation volume reference value, adjusts the circulation amount of the alkaline liquid of the multi-channel alkaline hydrogen production system and changes the gas purity of the multi-channel alkaline hydrogen production system by controlling the working states of the controllable channels on the two ends of the alkaline liquid circulating pump.

Abstract: A method, computer program product, and a system where a processor(s) determines generates a cognitive user profile representing patterns of usage of each of a plurality of users of the transportation resource sharing system, a cognitive resource profile for each resource of the plurality of resources, a cognitive route profile for each route traversed by at least one resource of the plurality of resources, and a cognitive station profile for each station of the plurality of stations. The processor(s) assigns one or more specific resources of the plurality of resources to one or more specific users of the plurality of users and the one or more specific resources of the plurality of resources to one or more specific stations of the plurality of stations.

Abstract: A method may include obtaining feature data associated with a set of feature inputs of a machine learning model, determining a probability that a subset of the feature data is concurrently used as the set of feature inputs for the machine learning model, and storing the subset of the feature data on a same cache node or server of a plurality of cache nodes or servers based on the probability.

Abstract: A machine learning system includes a training platform and an inference platform, where the inference platform is coupled to receive the output of the training platform. Based upon an updating of hyperparameters in the training platform, an optimized inference model is configured to be deployed to the inference platform from the training platform. The optimized inference model is further optimized in the inference platform by using an observation difference between a client observation and a prediction response to update the optimized inference model. The updated optimized inference model is used to provide a prediction response to a client.

Abstract: A direct-current coupling hydrogen production system includes at least one electricity generation system and multiple hydrogen production electrolyzer systems. The electricity generation system includes: a controller, N renewable energy systems, multiple conversion systems and a power switching unit. The power switching unit includes N input ports and M output ports. The controller is configured to control the power switching unit to supply the multiple hydrogen production electrolyzer systems through its output ports with electrical energy received through its input ports, or is configured to control the power switching unit to collect electrical energy received through its input ports and to supply the multiple hydrogen production electrolyzer systems through its output ports respectively corresponding to the hydrogen production electrolyzer systems with the collected electrical energy.

Abstract: Cache memory requirements between normal and peak operation may vary by two orders of magnitude or more. A cache memory management system for multi-tenant computing environments monitors memory requests and uses a pattern matching classifier to generate patterns which are then delivered to a neural network. The neural network is trained to predict near-future cache memory performance based on the current memory access patterns. An optimizer allocates cache memory among the tenants to ensure that each tenant has sufficient memory to meet its required service levels while avoiding the need to provision the computing environment with worst-case scenario levels of cache memory. System resources are preserved while maintaining required performance levels.

Abstract: A method, system, and computer program product for dynamically scheduling machine learning inference jobs receive or determine a plurality of performance profiles associated with a plurality of system resources, wherein each performance profile is associated with a machine learning model; receive a request for system resources for an inference job associated with the machine learning model; determine a system resource of the plurality of system resources for processing the inference job associated with the machine learning model based on the plurality of performance profiles and a quality of service requirement associated with the inference job; assign the system resource to the inference job for processing the inference job; receive result data associated with processing of the inference job with the system resource; and update based on the result data, a performance profile of the plurality of the performance profiles associated with the system resource and the machine learning model.

Abstract: Performing an operation comprising transforming an input dataset to a predefined format, extracting, from the transformed dataset, a plurality of features describing the transformed dataset, and generating, by a machine learning (ML) algorithm executing on a processor and based on an ML model, a plurality of rules for modifying the transformed dataset to conform with a first data model.

Abstract: The present application provides a display panel including a substrate, a thin film transistor layer, an electrode layer, a pixel definition layer, and a light-emitting function layer. layer. An orthographic projection area of the electrode on the substrate is D, an orthographic projection area of an area connecting the light-emitting function portion and the electrode on the substrate is P, and a reflection compensation parameter of the pixel is D/P. A display area comprises a first display area and a second display area. The reflection compensation parameter of the pixel in the first display area is less than the reflection compensation parameter of the pixel in the second display area.

Abstract: A method for manufacturing a flex-tolerant three-dimensional printed antenna suitable for use in an electronic device provides a three-dimensional printed antenna with base layer, radiation layer, through holes, and feeder. The radiation layer includes a first radiation region, at least one second radiation region, and a feed end. A region between the first and second radiation regions is defined as a bent region. The radiation layer is formed by a screen-printing plate by a planar printing process. The through holes on the bent region form a line for bending. The feeder is electrically connected to the feed end. The second radiation region is canted from the bending line with respect to the first radiation region to form the three-dimensional printed antenna. The three-dimensional printed antenna and an electronic device are also disclosed.

Abstract: A method for manufacturing a bending-tolerant multilayered flexible circuit board (FPC) suitable for use in a disposable biosensor chip includes manufacturing and bending a single-sided FPC. The single-sided FPC includes a base layer, a wiring layer, and through holes. The wiring layer includes first and second wiring areas. A first bending area is formed between each first wiring area and the corresponding second wiring area, the through holes forming an easy bending line. The second wiring area is bent relative to the first wiring area along the bending line to obtain the multilayered FPC.