7results about How to "Improve learning accuracy" patented technology

The embodiment of the invention provides a federated learning model parameter updating method, system and device, and the method comprises the steps: receiving local model parameters sent by N clients under the condition that the jth round of model parameter updating is carried out on federated learning, and obtaining N groups of local model parameters of the jth round; clustering the N groups of local model parameters of the jth round through K clustering center model parameters of the (j-1) th round to obtain K clusters; updating the K clusters according to the distance between the local model parameters of the jth round and the cluster center model parameters of the (j-1) th round of each group in the clusters to obtain updated K clusters, and obtaining the K cluster center model parameters of the jth round and the global model parameters of the jth round through the updated K clusters. Through the method and the device, the problem of low federal learning accuracy caused by incapability of eliminating error model parameters transmitted by an abnormal client in related technologies is solved.

The present application belongs to the field of feature extraction in machine learning, in particular to a device damage feature screening method based on information entropy and Person correlation coefficient. First, according to the principle of power law, the method of fuzzy mutual information is used to select strong correlation features, in order to reduce redundancy, by referring to matrix transformation, the Person correlation coefficient is used to calculate the correlation between strong correlation features, and the redundant features with strong correlation are removed, finally, in order to improve the complementarity, whether there is strong correlation based on combination in the remaining other features is considered by using joint adjustment mutual information calculation. Compared with the prior art, the present application not only considers the correlation between features, but also reduces the redundancy between features, and also considers the complementarity between features from the joint feature angle, thereby maximizing the accuracy of feature selection and improving the efficiency of machine learning method.

The present invention discloses a tooth profile learning device for a key processing machine and a key processing machine, comprising a cutter head probe for obtaining tooth profile data of an original key and a first transmission component; the first transmission component comprises a first shaft, a first transmission mechanism and a first motor, the first motor drives the first shaft to rotate through the first transmission mechanism, one end of the cutter head probe is connected to the first shaft, and the other end comprises a contact portion for obtaining tooth profile data of the original key, when the first shaft drives the cutter head probe, the contact portion moves back and forth between a position matching the original key and a position away from the original key, compared with the current contact-type tooth profile learning device, no additional reference calibration structure is required, and no additional manual operation is required, so the use is more convenient and efficient.

The present application discloses an offline reinforcement learning training method, an action prediction method, a device and a medium, wherein the method comprises: obtaining a first static data set of a robot controlled by a model to be trained; the first static data set comprises a plurality of historical running trajectories of the robot; the network of the model to be trained comprises a value function network, a diffusion model network and an inverse dynamics network; training the value function network based on the first static data set; generating performance labels corresponding to each trajectory segment of the historical running trajectory based on the trained value function network; training the diffusion model network and the inverse dynamics network based on each trajectory segment and each performance label. The method and device provided by the present application improve the learning accuracy of the diffusion model, and improve the training efficiency and prediction accuracy of the model to be trained.

The invention relates to the technical field of oilfield development, and discloses a construction method and application of a mineral content well logging intelligent interpretation model, reservoirs in a target rock area are classified by using a ternary comprehensive classification method composed of lithology, porosity and oil-bearing level, interference of the lithology, porosity and oil-bearing level on mineral content well logging interpretation can be reduced, and the interpretation accuracy is improved. The model learning precision is improved; multi-wavelet decomposition and reconstruction are carried out on logging curve data, all effective information of an original logging curve is reserved, the redundant degree between wavelet components is reduced, and the logging curve difference caused by the non-mineral content is effectively reduced; and based on the idea of an integration algorithm, training the intelligent mineral content logging interpretation model of each reservoir by using an optimal machine learning algorithm corresponding to different types of reservoirs, and then integrating the intelligent mineral content logging interpretation models of different types of reservoirs to obtain the intelligent mineral content logging interpretation model of the target rock area. The overall accurate mineral content identification of the target rock area can be realized.

The present application provides an intelligent driving path learning method, an automatic cruise method, related equipment and a vehicle, wherein the method includes: obtaining a preset driving path; according to the driving path and the navigation map, using the first path information on the driving path collected by the vehicle body sensor, and the second path information on the driving path obtained based on the navigation map to complete the path learning, wherein the first path information and the second path information include different path information, and the first path information and the second path information jointly determine all the path information required for the path learning. The path learning method provided by the present application uses the vehicle sensor to collect path information, and assists in collecting path information based on the navigation map, which reduces the amount of data collected by the vehicle sensor, improves the efficiency of path learning, and can also supplement some information that cannot be collected by the vehicle sensor, thereby improving the accuracy of path learning.

A motor torque learning device that learns an optimal value of an instruction torque of a motor in order to maximize a performance index value and stop rotation of a crank at a target stop position, the motor torque learning device including: a storage device that stores an environment model that takes a first engine speed and a first crank position during an internal combustion engine stop process and an instruction torque as inputs and outputs a changed second engine speed and a second crank position corresponding to application of the instruction torque as input to the crank and a performance index value as a reward obtained with application of the instruction torque; and a policy learning unit that learns a relationship of the optimal value of the instruction torque for maximizing the reward with respect to the first engine speed and the first crank position by changing the instruction torque and repeatedly performing an operation of obtaining an output of the environment model with respect to inputs of the first engine speed, the first crank position, and the instruction torque.