Abstract: Embodiments of the present disclosure relate to a method, an electronic device, and a computer program product for training a data classification model. The method includes generating a first training rule based on probabilities of classifying a plurality of sample data into corresponding classes by a data classification model. The method also includes generating a second training rule based on relevances of the plurality of sample data to the corresponding classes. In addition, the method also includes training the data classification model using the first training rule and the second training rule. With this method, a data classification model is trained, so that the data classification accuracy of the data classification model and the robustness to noise can be improved.

Abstract: Provided are an interaction method and apparatus in a live streaming room, and a device and a storage medium. The method comprises: in response to a trigger operation for a preset duet entry on a live streaming room page, jumping from the live streaming room page to a camera capture page, wherein a host portrait cutout corresponding to the live streaming room page is displayed on the camera capture page and a capture control is provided on the camera capture page acquiring multimedia information captured by a camera, in response to a trigger operation for the capture control on the camera capture page, wherein the multimedia information comprises an image or a video; and synthesizing the multimedia information with the host portrait cutout to obtain a duet multimedia product.

Abstract: A method in one embodiment includes receiving, at an edge device, new data for training a model, the edge device having stored distilled data used to represent historical data to train the model, the historical data being stored in a remote device, and the amount of the historical data being greater than the amount of the distilled data. The method further includes training the model based on the new data and the distilled data. With the data processing solution of this embodiment, the model can be trained at the edge device with fewer storage resources based on the distilled data, thereby achieving higher model accuracy.

Abstract: This application relates to processes and intermediates for the preparation of the PD-1/PD-L1 inhibitor (R)-1-((7-cyano-2-(3?-((2-(difluoromethyl)-7-((3-hydroxypyrrolidin-1-yl)methyl)pyrido[3,2-d]pyrimidin-4-yl)amino)-2,2?-dimethyl-[1,1?-biphenyl]-3-yl)benzo[d]oxazol-5-yl)methyl)piperidine-4-carboxylic acid, and salts and crystalline forms thereof, where the PD-1/PD-L1 inhibitor and solid forms and salt forms thereof are useful in the treatment of various diseases including infectious diseases and cancer.

Abstract: Embodiments of the present disclosure provide a method, an electronic device, and a computer program product for information processing. In an information processing method, based on multiple weights corresponding to multiple words in text, a computing device determines a target object associated with the text among predetermined multiple objects, and also determines, among the multiple words, a set of key words with respect to the determination of the target object. Next, the computing device determines, among the set of key words, a set of target words related to a text topic of the text. Then, the computing device outputs the set of target words and an identifier of the target object in an associated manner. In this way, the credibility of the target object associated with the text that is determined by the information processing method is improved, thereby improving the user experience of the information processing method.

Abstract: Embodiments of the present disclosure relate to a computer-implemented method, device, and computer program product. The method includes: determining, based on a set of user behavior data over a first time period, a set of behavioral features for a user behavior over the first time period. The method further includes: determining, based on the set of behavioral features, a set of anomaly scores for the user behavior according to an anomaly detection model. The anomaly detection model is trained based on user behavior data over a second time period. The method further includes: updating a previously determined anomaly score threshold based on comparison of the determined set of anomaly scores with the anomaly score threshold. The anomaly score threshold is used for indicating whether the user behavior is anomalous. By continuously updating the anomaly score threshold, the solution uses the updated anomaly score threshold for anomaly detection of user behavior.

Abstract: A method in an illustrative embodiment of the present disclosure includes determining, utilizing a first diagnosis model deployed in a storage system, whether a cause of a fault belongs to environmental factors. The method further includes determining, responsive to determining that the cause of the fault belongs to the environmental factors, whether the fault can be solved locally in the storage system. The method further includes sending, responsive to determining that the fault cannot be solved locally in the storage system, the fault to a second diagnosis model, wherein the first diagnosis model is obtained by distilling the second diagnosis model. According to the method for fault diagnosis of the present disclosure, particular faults can be diagnosed and solved locally in a storage system, so that the workload of a customer support team of the storage system in a cloud can be reduced.

Abstract: Embodiments of the present disclosure provide a method, an electronic device, and a medium for bug classification. The method includes determining, based on description information of a bug generated during product testing, classification information of the bug through at least one trained computing model; presenting the classification information of the bug; determining, based on user interaction for the presented classification information, whether performance of the at least one computing model satisfies a predetermined condition; and determining that the at least one computing model needs to be retrained in response to determining that the performance of the at least one computing model does not satisfy the predetermined condition. In this way, automatic classification of the bug is realized, and the computing model can be dynamically adjusted by retraining, so as to ensure accuracy of the automatic classification and improve efficiency of bug processing.

Abstract: Embodiments of the present disclosure provide a method, an electronic device, and a computer program product for detecting model performance. The method may include acquiring a prediction result of an input feature using a target model to determine a confidence of the prediction result. The method may further include reconstructing the input feature using a self-coding model to determine a reconstruction error, the reconstruction error being a difference between the input feature before being reconstructed by the self-coding model and the input feature after being reconstructed by the self-coding model. In addition, the method may include determining a detection result of the target model at least based on a comparison between the confidence and a first threshold and a comparison between the reconstruction error and a second threshold.

Abstract: Embodiments of the present disclosure relate to a method, a device, and a computer program product for model arrangement. The method includes determining a target model for processing data. The method further includes dividing the target model into a plurality of modules that implement different tasks. The method further includes determining a quantity of parameters of a target module in the plurality of modules and a size of transmission data related to the target module. The method further includes determining an arrangement position of the target module based on the quantity and the size. With this method, the amount of data transmitted can be minimized, and the computing time and the presentation time of information presented to users can be reduced, thereby improving the user experience.

Abstract: A touch substrate manufacturing method, for manufacturing a target touch substrate to be assembled with a display substrate, is provided. A size of an active display region of the display substrate is a first preset size. The method includes: providing a touch substrate to be processed having a second preset size, the second preset size being greater than the first preset size; and cutting, in at least one cutting direction, the touch substrate to be processed to obtain the target touch substrate, the cutting direction being parallel to an extending direction of a touch channel on the touch substrate to be processed. A manufacturing method for touch module, a touch substrate and a touch module are also provided.

Abstract: A method in an illustrative embodiment of the present disclosure includes determining, utilizing a first diagnosis model deployed in a storage system, whether a cause of a fault belongs to environmental factors. The method further includes determining, responsive to determining that the cause of the fault belongs to the environmental factors, whether the fault can be solved locally in the storage system. The method further includes sending, responsive to determining that the fault cannot be solved locally in the storage system, the fault to a second diagnosis model, wherein the first diagnosis model is obtained by distilling the second diagnosis model. According to the method for fault diagnosis of the present disclosure, particular faults can be diagnosed and solved locally in a storage system, so that the workload of a customer support team of the storage system in a cloud can be reduced.

Abstract: An apparatus comprises a processing device configured to train first and second machine learning models utilizing a first training dataset comprising inputs each associated with a class label of one of a set of classes and a second training dataset comprising distilled representations of the two or more classes, and to identify candidate adversarial example inputs utilizing the trained first and second machine learning models. The processing device is further configured to determine whether the candidate adversarial example inputs are true positive adversarial example inputs based on a confidence-aware clustering and to generate an updated first training dataset comprising corrected class labels for the true positive adversarial example inputs and an updated second training dataset comprising updated distilled representations determined utilizing the corrected class labels.

Abstract: Embodiments of the present disclosure relate to a method, an electronic device, and a computer program product for retrieving service requests. The method includes determining a hash value of a service request based on the service request. The method further includes determining request pairs based on correlations between the hash value of the service request and a plurality of hash values of a plurality of historical service requests. The method further includes determining a semantic correlation between the service request and the historical service request in each of the plurality of request pairs. The method further includes determining, based on the determined semantic correlation between each request pair, a probability indicating that the service request and the historical service request in the request pair use the same solution. The method further includes determining a retrieved historical service request based on the probability.

Abstract: Disclosed is a solid-state transformer having an uninterrupted operation ability under an AC/DC fault, in which bridge arms of a hybrid modular multilevel converter include half-bridge submodules and full-bridge submodules. The half-bridge submodules and the full-bridge submodules are connected with input ends of isolated dual-active-bridge converters via DC capacitors of the half-bridge submodules and full-bridge submodules; output ends of the isolated dual-active-bridge converters are connected in parallel to form a low-voltage DC bus; and a three-phase full-bridge inverter is connected to the low-voltage DC bus. The solid-state transformer may be provided with four ports including a medium-voltage AC port, a medium-voltage DC port, a low-voltage DC port and a low-voltage AC port which are beneficial to the interconnection of multi-voltage-level and multi-form AC/DC hybrid distribution grids. The solid-state transformer has the uninterrupted operation ability under the AC/DC fault.

Abstract: The invention relates to an island converter overload limit method and device for a bipolar flexible DC transmission system, and belongs to the field of DC transmission. During stable island operation, both bipolar converters adopt a voltage-frequency droop control strategy; and once it is detected that one converter is overloaded, the control mode of the overloaded converter is automatically switched to active-wattless power control from voltage-frequency droop control to fulfill an overload limit function. The method effectively avoids an overload of the bipolar flexible DC transmission system under island operation, effectively maintains the AC voltage and frequency stable, and has important guiding significance for applying island systems to DC power grids.

Abstract: Disclosed are a control method and system for a parallel converter system, including a common AC voltage controller and a separate current controller for each converter; the AC voltage controller generates an active current, a wattless current, and a system voltage phase reference value and, by means of communication, sends same to the current controller of each converter in island control mode to be a converter control signal; the active and wattless current outputted by each converter, along with the corresponding current reference value, accomplish the objective of the parallel converter system collectively controlling the amplitude and frequency of an AC bus voltage. The method and system are highly reliable and effectively ensure the synchronization of a plurality of converters.

Abstract: In a solid-state transformer having an uninterrupted operation ability under an AC/DC fault, bridge arms of a hybrid modular multilevel converter comprise half-bridge submodules and full-bridge submodules, the half-bridge submodules and the full-bridge submodules are interconnected with input ends of the isolated dual-active-bridge converters via DC capacitors of the submodules, output ends of the plurality of isolated dual-active-bridge converters are connected in parallel to form a low-voltage DC bus, and the three-phase full-bridge inverter is connected to the low-voltage DC bus. The solid-state transformer provided by the present invention may be provided with four ports including a medium-voltage AC port, a medium-voltage DC port, a low-voltage DC port and a low-voltage AC port which are beneficial to the interconnection of multi-voltage-grade and multi-form AC/DC hybrid distribution grids; and the solid-state transformer has the uninterrupted operation ability under the AC/DC fault.

Abstract: Systems and methods for generating telecast forecasts is provided. An automated forecasting system uses machine learning-driven a forecast model for generating forecast for various telecasts varying periods of time. Estimate values that are used to generate the forecasts may be determined based on deriving trends and correlations from telecasts data using machine learning. The forecasting system may compare estimate values and actual values associated with the various telecasts and subsequently update the forecast model based on the comparison. The forecast model may be displayed on an electronic device of a client electronic device and may be updated or influenced by telecast providers.

Abstract: A DC voltage coordination control method is applied to a multi-terminal VSC-HVDC power transmission system or a VSC-HVDC power grid system. When a lower-level control is used in the active power control mode, the active power reference value of the converter is generated according to the converter DC voltage and the converter DC voltage active power curve; it also includes an upper-level control to adjust the control mode and active power setting value of the lower-level control to ensure the steady-state performance of the system. This method has low dependence on communication, avoids the problem that the fixed slope and margin cannot meet all operating conditions, and is suitable for large-scale VSC-HVDC systems, and has scalability.