Abstract: Systems, methods, and computer program products that obtain a plurality of features associated with a plurality of samples and a plurality of labels for the plurality of samples; generate a plurality of first predictions for the plurality of samples with a first machine learning model; generate a plurality of second predictions for the plurality of samples with a second machine learning model; generate, based on the plurality of first predictions, the plurality of second predictions, the plurality of labels, and a plurality of groups of samples of the plurality of samples; determine, based on the plurality of groups of samples, a first success rate associated with the first machine learning model and a second success rate associated with the second machine learning model; and identify, based on the first success rate and the second success rate, a weak point in the machine learning first model or the second model.

Abstract: Provided in the present disclosure are a memory and apparatus for managing a memory snapshot, a device, and a medium. The method includes: when a memory snapshot generation command is acquired, a virtual machine is suspended; a two-level page table of the virtual machine is copied to obtain a first snapshot page table, and a read-only identifier is set for each of page table entries of the two-level page table; and memory pages corresponding to page table entries of the first snapshot page table are acquired by means of the page table entries of the first snapshot page table, a reference count of each of the memory pages is increased, and the virtual machine is restored from a suspended state to a running state.

Abstract: The present teaching relates to a fraud detecting system and method for providing protection against fraudulent advertisement requests. Upon receiving a request for an advertisement, the system extracts an identifier, associated with a source from which the request originates, included in the request. The system determines whether the extracted identifier is included in a list of designated identifiers, and when the identifier is included in the list, the system denies the request for the advertisement. When the identifier is not included in the list of designated identifiers, the system provides the advertisement in response to the request, and extracts a set of features from the request and other requests that originate from the source to determine whether the identifier associated with the source is to be included in the list of designated identifiers based on the set of features in accordance with one or more models.

Abstract: An Intelligent Electronic Device (IED) for measuring a zero-crossing based electrical parameter when there is power quality disturbance in the power supply network. More specifically, the IED includes at least one sensor configured for sensing at least one electrical parameter of electrical power distributed from an electrical distribution system to a load, at least one analog-to-digital converter coupled to the at least one sensor and configured for converting an analog signal output from the at least one sensor to digital data, at least one processing module coupled to the at least one analog-to-digital converter, the at least one processing module is configured to receive the digital data, determine that the IED enters loss of synchronization according to the received digital data, measure a zero-crossing based electrical parameter of the digital data according to the cycle of the digital signal immediately before loss of synchronization.

Abstract: Methods, systems, and computer program products are provided for embedding learning to provide uniformity and orthogonality of embeddings. A method may include receiving a dataset that includes a plurality of data points including a first plurality of data points having a first classification and a second plurality of data points having a second classification, generating a first normalized class mean vector of the first plurality of data instances having the first classification, generating a second normalized class mean vector of the second plurality of data instances having the second classification, performing a class rectification operation on the first plurality of data instances having the first classification and the second plurality of data instances having a second classification, and generating embeddings of the dataset based on original embedding space projections of the dataset.

Abstract: The present disclosure provides a honeycomb catalyst for catalytic oxidative degradation of VOCs prepared by an ultrasonic double-atomization process.

Abstract: Embodiments of this application relate to the field of short-range communications technologies, and provide a method for automatically switching a Bluetooth audio coding scheme and an electronic device, to improve audio experience of using a Bluetooth electronic device by a user, and meet requirements of the user on sound quality and a latency. A specific solution is as follows: When a first device sends a Bluetooth audio data packet to a second device, and when a first application is running, the audio data packet sent by the first device to the second device is encoded by using a first Bluetooth coding scheme. When a second application is running, the audio data packet sent by the first device to the second device is encoded by using a second Bluetooth coding scheme. The Bluetooth coding scheme includes a high sound quality mode, a standard mode, or a low latency mode. The first Bluetooth coding scheme is different from the second Bluetooth coding scheme.

Abstract: A method performed by a server computer is disclosed. The method comprises generating a binary compositional code matrix from an input matrix. The binary compositional code matrix is then converted into an integer code matrix. Each row of the integer code matrix is input into a decoder, including plurality of codebooks, to output a summed vector for each row. The method then includes inputting a derivative of each summed vector into a downstream machine learning model to output a prediction.

Abstract: The system and method may assess the merchant risk level on a more continuous scale rather than a binary categorization. It may produce a continuous risk score proportional to the likelihood of a merchant being risky, effectively addressing the issue of shades of gray encountered by the traditional blacklisting approach. The continuous risk score feature provides greater flexibility as it allows the payment network to make dynamic pricing decisions (known as interchange optimization) based on the merchant risk level. Using collective intelligence from transactions across the payment network, the system and method may be able to assess the merchant risk level with high accuracy. The system and method may be particularly beneficial to small merchants with low transaction volume as even a few fraudulent transactions can easily put them in the high-risk merchant category. Further, the system and method may help payment processing networks make better decision on cross-border transactions.

Abstract: Provided are an audio and video transcoding method performed by a computer device. The method includes: obtaining first multimedia data in a first format; processing the first multimedia data in the first format into intermediate data through a first transcoding operation; processing the intermediate data into at least two pieces of second multimedia data in second formats through at least two second transcoding operations, the first transcoding operation and the at least two second transcoding operations being operations performed based on that a media bus provides a data communication channel; and outputting the at least two pieces of second multimedia data in the second formats.

Abstract: Embodiments of this application provide a Bluetooth communication method and system. The method includes: A first electronic device to be added to a BLE network may actively send a request for joining the BLE network, and a second electronic device in the BLE network may prompt, in response to the request of the first electronic device, a third electronic device that currently, there is the first electronic device to be added to the BLE network. The third electronic device may display a pop-up prompt, to notify a user that currently, there is the first electronic device to be added to the BLE network. After the user agrees, the third electronic device prompts the second electronic device to configure the BLE network for the first electronic device.

Abstract: A microfluidic apparatus (100) can include a PCB (110), a biological chip (120) overlying the PCB (110), and a microfluidic housing (130) overlying the biological chip (120) and the PCB (110). The microfluidic apparatus (100) also has a first adhesive layer (141) attaching the microfluidic housing (130) to the biological chip (120) and a second adhesive layer (142) attaching the microfluidic housing (130) to the PCB (110). The second adhesive layer (142) is thicker than the first adhesive layer (141). The first adhesive layer (141) comprises a first adhesive material, and the second adhesive layer (142) comprises a second adhesive material.

Abstract: Overlapping data streams and sensor transitions is described. In accordance with the described techniques, a first data stream of first analyte measurements is received, at a computing device and during an overlap period, from a first analyte sensor worn by a user and a second data stream of second analyte measurements is received, at the computing device and during the overlap period, from a second analyte sensor worn by the user. The first data stream and the second data stream are received concurrently during the overlap period. A combined analyte measurement is generated based on the first analyte measurements and the second analyte measurements. In one or more implementations, the combined analyte measurement is generated as a weighted average of the first analyte measurements and the second analyte measurements. The combined analyte measurement is then output on a display of the computing device.

Abstract: A system and method of manufacturing an aluminum fuel cell cradle includes providing a negative cast mold having cavities to form the cradle having a length greater than a width to define a longitudinal axis and providing a feeding mechanism disposed about the mold and in fluid communication with the cavities thereof. The feeding mechanism includes primary risers connected to and in fluid communication with the cavities. The method further includes melting a first metallic material to define a molten metallic material, moving the mold to a vertical orientation about the longitudinal axis while feeding molten metallic material through the runner to the cavities, and moving the cast mold to a horizontal orientation. A second solidification time in the risers is greater than a first solidification time in the mold such that shrinkage of the solidified metallic material occurs in the risers away from the mold.

Abstract: An apparatus (100) including multiple biological chips (110,120) includes a substrate (101), a first adhesive layer (134) disposed on the substrate (101), a first biological chip (110) and a second biological chip (120) disposed on the first adhesive layer (134) and attached to the substrate (101) by the adhesive layer (134). The apparatus (100) further includes a filler (130) disposed between the first biological chip (110) and the second biological chip (120). The filler (130) includes a second adhesive layer (135) extending between a side surface (114) of the first biological chip (110) and a side surface (124) of the second biological chip (120), the second adhesive layer (135) attaching the first biological chip (110) to the second biological chip (120). The filler (130) also includes a surface layer (132) disposed over the second adhesive layer (135).

Abstract: A voltage detection method includes: outputting a corresponding one of voltages to be an input voltage according to a clock signal, a first detection signal, a reset signal and multiple first bits; generating the reset signal according to the clock signal and multiple currents, wherein the voltages and the currents are generated based on a power supply voltage; comparing the input voltage with a reference voltage to generate a second detection signal, and generating the first detection signal according to the second detection signal and an enable signal; and adjusting the first bits by a digital circuit according to the second detection signal during a trimming phase to determine the first bits, and outputting the first bits by the digital circuit and resetting the digital circuit according to the first detection signal during a voltage detection phase.

Abstract: A system and method of manufacturing an aluminum fuel cell cradle includes providing a negative cast mold having cavities to form the cradle and providing a feeding mechanism disposed about the mold and in fluid communication with the cavities thereof. The feeding mechanism includes a plurality of primary risers connected to and in fluid communication with cavities. The method further includes melting a first metallic material to define a molten metallic material, and moving the mold to a vertical casting orientation about a rotational axis, while feeding molten metallic material through the runner to the cavities, and cooling the molten metallic material to define a solidified metallic material. A second solidification time in the risers is greater than a first solidification time in the mold such that shrinkage of the solidified metallic material occurs in the risers away from the mold.

Abstract: A power control device and method, the method including obtaining, by a terminal device, class of device information of a peer device, determining, by the terminal device, based on the class of device information, a class of device of the peer device, and performing at least one of sending, by the terminal device, in response to the terminal device determining that the class of device of the peer device is a non-preset class of device, data to the peer device at a first transmit power, or sending, by the terminal device, data to the peer device at a second transmit power—in response to the terminal device determining that the class of device of the peer device is a preset class of device, —where the terminal device establishes a Bluetooth connection to the peer device, and the first transmit power is less than the second transmit power.

Abstract: A casting die includes a first ingate defining a first annular chamber and a second ingate defining a second annular chamber and spaced from the first ingate. The die includes a first end ring gate concentric with the first ingate and defining a first end ring chamber in fluid communication with the first annular chamber, and a second end ring gate concentric with the second ingate and defining a second end ring chamber in fluid communication with the second annular chamber. The die includes runners interconnecting the first and second end ring gates. Each of the runners defines a conductor bar channel in fluid communication with the first and second end ring chambers. The die includes an overflow gate disposed between the first and second ingates, wherein the overflow gate defines an annular overflow chamber in fluid communication with the conductor.