Abstract: The invention provides a method for constructing a target prediction model in a multicenter small sample scenario and a prediction method. In combination with the idea of transfer learning, a training set of a new node is predicted by directly using knowledge of a trained node, and a prediction error sample is used to reflect a difference between the new node and the trained node. The difference is used as supplementary knowledge. Model knowledge of the new node is quickly acquired, to avoid training of a new node from scratch each time. Finally, parallel integration of incremental subclassifiers is implemented by using a ridge regression method, so that the deployment time and costs are greatly reduced. The generalization of models is ensured through sharing of historical knowledge and a knowledge discarding mechanism, good classification effect can also be achieved for a node with a small sample size.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed for teacher-free self-feature distillation training of machine-learning (ML) models. An example apparatus includes at least one memory, instructions, and processor circuitry to at least one of execute or instantiate the instructions to perform a first comparison of (i) a first group of a first set of feature channels (FCs) of an ML model and (ii) a second group of the first set, perform a second comparison of (iii) a first group of a second set of FCs of the ML model and one of (iv) a third group of the first set or a first group of a third set of FCs of the ML model, adjust parameter(s) of the ML model based on the first and/or second comparisons, and, in response to an error value satisfying a threshold, deploy the ML model to execute a workload based on the parameter(s).

Abstract: The present invention provides a multi-segment bendable joint, comprising: a jointer, having a rod portion, a first ball head and a second ball head connected at two ends of the rod portion; a first connector, at opposite ends thereof, with a first connecting portion and a first joint slot for receiving the first ball head therein, respectively; and a second connector, at opposite ends thereof, with a second connecting portion and a second joint slot for receiving the second ball head therein, respectively. When one of the first and second connectors is linked to a wrench body and the other acts as a socket, two-segment bending between the socket and the wrench body is allowed, resulting in larger bending angle between the socket and the wrench body compared to conventional products.

Abstract: A lattice-based proxy signature method, apparatus and device, a lattice-based proxy signature verification method, apparatus and device, and a storage medium. Polynomials are randomly selected in rings to calculate public and private keys of nodes, and the magnitudes of proxy public and private keys are the same as the magnitudes of public and private keys of an original signer. Therefore, compared with existing proxy signature schemes, the present application has smaller lengths of public and private keys and higher storage efficiency. Proxy signature information generated in the present application shows a signature of the original signer and also shows a signature of a proxy signer. Once a proxy signature is created, the proxy signature cannot be repudiated by the proxy signer, and has strong non-repudiation and strong unforgeability. The proxy signature method has the advantage of resisting quantum computer attack.

Abstract: A computer-implemented method is provided that includes transmitting, by a master node to a plurality of computing nodes, definition information about an initial medical validation model (410): performing, by the master node, a federated learning process together with the plurality of computing nodes (420), to jointly train the initial medical validation model using respective processed local training datasets available at the plurality of computing nodes, the respective local training datasets being processed by the plurality of computing nodes based on the definition information; and determining, by the master node, a final medical validation model based on a result of the federated learning process (430). Through the solution, by means of federated learning, it addresses the data security and privacy concerns from local sites owning.

Abstract: Technology providing in-memory neural network protection can include a memory to store a neural network, and a processor executing instructions to generate a neural network memory structure having a plurality of memory blocks in the memory, scatter the neural network among the plurality of memory blocks based on a randomized memory storage pattern, and reshuffle the neural network among the plurality of memory blocks based on a neural network memory access pattern. Scattering the neural network model can include dividing each layer of the neural network into a plurality of chunks, for each layer, selecting, for each chunk of the plurality of chunks, one of the plurality of memory blocks based on the randomized memory storage pattern, and storing each chunk in the respective selected memory block. The plurality of memory blocks can be organized into a groups of memory blocks and be divided between stack space and heap space.

Abstract: An accelerator includes a memory, a compute zone to receive an encrypted workload downloaded from a tenant application running in a virtual machine on a host computing system attached to the accelerator, and a processor subsystem to execute a cryptographic key exchange protocol with the tenant application to derive a session key for the compute zone and to program the session key into the compute zone. The compute zone is to decrypt the encrypted workload using the session key, receive an encrypted data stream from the tenant application, decrypt the encrypted data stream using the session key, and process the decrypted data stream by executing the workload to produce metadata.

Abstract: A method and an apparatus for automatic cooking control based on an intelligent cooking device, which is applied to an intelligent cooking device, the intelligent cooking device including an intelligent cookware and an intelligent stove, the method including: collecting a present ingredient parameter of a target ingredient and a present cooking temperature, the present ingredient parameter comprising at least an ingredient identification; acquiring a present cooking gear; generating a first cooking control parameter according to the present ingredient parameter, the present cooking temperature and the present cooking gear, the first cooking control parameter comprising at least a first stove control parameter and/or a first cookware control parameter; and controlling the intelligent cooking device to cook the target ingredient according to the first cooking control parameter.

Abstract: A method and an apparatus for recipe preparation based on ingredient-related data, the method includes: acquiring ingredient-related data of a target combination of ingredients uploaded by a user, the ingredient-related data comprising one or more of an ingredient type, a cooking process, a cooking cuisine, a finished cooking product and a cooking performance; analyzing a pending optimization situation corresponding to the target combination of ingredients according to the ingredient-related data; and generating a cooking recipe corresponding to the target combination of ingredients according to the pending optimization situation and the ingredient-related data.

Abstract: Systems, apparatuses and methods include technology that identifies data that is to be transferred from a first device to a second device. The technology classifies the data into a category from a plurality of categories, selects a compression scheme from a plurality of compression schemes based on the category and compresses the data based on the compression scheme.

Abstract: It relates to an apparatus, a device, a method, and a computer program for generating test cases for a verification of hardware instructions of a hardware device in a hypervisor. The apparatus comprises circuitry configured to generate a transition table based on a specification of the hardware device. The transition table comprises a plurality of entries. Each entry represents a change of a state of the hardware device in response to an event. The circuitry is configured to determine entries of the transition table that are equivalent. The circuitry is configured to generate a plurality of test cases based on the entries of the transition table. At least one entry of the transition table is omitted in the generation of the test cases due to being equivalent to another entry of the transition table.

Abstract: A control apparatus (10), control device, control method and computer program for controlling one or more parameters of a hypervisor (100) and an apparatus, device, method, and computer program for a virtual machine (200). The control apparatus (10) comprises circuitry configured to obtain information on respective performance targets of two or more virtual machines (200) being hosted by the hypervisor (100). The circuitry is configured to set the one or more parameters of the hypervisor (100) to one or more initial values. The circuitry is configured to obtain respective results of a benchmark being run in the two or more virtual machines (200), the results of the benchmark indicating a performance of the respective virtual machines (200) with respect to the respective performance targets, with the results of the benchmark being affected by the one or more parameters.

Abstract: Examples relate to an apparatus, a device, a method, and a computer program for generating a test specification for testing software code of a function under test. The apparatus for generating the test specification for testing software code of a function under test comprises circuitry configured to extract a plurality of symbols from the software code of the function under test, generate a plurality of test vectors with corresponding sets of expected results for the function under test based on the plurality of symbols, and generate a test specification based on the plurality of test vectors and the corresponding sets of expected results.

Abstract: Disclosed are 6-carbamate substituted heteroaryl ring derivatives and a preparation method therefor. The present application particularly relates to a compound as shown in formula (II) and a pharmaceutically acceptable salt thereof.

Abstract: The present invention relates to a thiophene compound and an application thereof. The present invention specifically relates to a compound represented by formula (II) and a pharmaceutically acceptable salt thereof, wherein L1 is —C(?O)— or —C(?O)—NH—; R1 is R1a or aa; ring A is bb or cc; R1a is C1-3 alkyl substituted with 1, 2 or 3 Ra; and R3 is independently H, halogen or C1-3 alkoxy. The compound represented by formula (II) and the pharmaceutically acceptable salt thereof can better inhibit the activity of LSD1, further shows significant inhibitory activity against NCI-H1417 cell proliferation, and has good pharmacokinetic properties (including good oral bioavailability, oral exposure, half-life, and clearance rate, etc.).

Abstract: Examples relate to an apparatus, a device, a method and a computer program for generating code for test cases for testing a function under test. The apparatus comprises circuitry configured to obtain a diagram representation of the function under test, select a plurality of symbols of interest from the diagram representation of the function under test, the symbols of interest being based on a pre-defined set of symbols of interest, and generate, for the symbols of interest, code of a plurality of test cases based on a pre-defined set of checks related to the pre-defined symbols of interest.

Abstract: The disclosure relates to content aware foveated ASW for low latency rendering. A device for graphics processing comprises processing resources and a pipeline at least partly implemented by the processing resources. The pipeline is to: render input data to generate a first frame; perform ASW on the first frame on a block basis based on content and focusing related information associated with the first frame to generate a second frame; and output the first frame and the second frame.

Abstract: The disclosure relates to tuning configuration parameters for graphics pipeline for better user experience. A device for graphics processing, comprising: hardware engines; a graphics pipeline at least partly implemented by the hardware engines; and a tuner, coupled to the hardware engines and the graphics pipeline, the tuner to: collect statuses of the device during runtime for a previous frame; determine configuration parameters based on the collected statuses, the configuration parameters associated with three-dimensional 3D rendering, pre-processing and video encoding of the graphics pipeline; and tune the graphics pipeline with the determined configuration parameters for processing a next frame.

Abstract: Various technologies relating to wireless sensor networks (WSNs) are disclosed, including, but not limited to, device onboarding and authentication, network association and synchronization, data logging and reporting, asset tracking, and automated flight state detection.

Abstract: The invention belongs to the field of organic synthesis, and specifically relates to the hexadecane tromethamine compound which has the following structure: and provides the synthesis method of the hexadecane tromethamine compound, and the application of the hexadecane tromethamine compound in anti-tumor and antifungal aspects. The invention fills the gaps in the synthesis process of the hexadecane tromethamine compound and the application of the hexadecane tromethamine compound in anti-tumor and antifungal aspects.