7 results about "Model selection" patented technology

The embodiment of the invention provides an AI-based universal smart home scheme automatic design method and device and electronic equipment. The method belongs to the technical field of artificial intelligence, and comprises the following steps: collecting a plane drawing related to intelligent design as an effective image sample, establishing convolutional layer parameter sharing and sparse connection for a neural network system, and reducing the size of a model by using a pooling layer; reducing effective identification of the plane drawing through Python language programming, intelligentlyadding alternative product icons of a database, and zooming to a proper size to obtain point location design; and obtaining an input demand set by a user at a client, and generating a general intelligent home design scheme according to confirmation of model selection of the user at the client. By adopting the scheme, the efficiency of smart home scheme design can be improved.

The invention provides an integrated pump station management method, device and equipment and a readable storage medium, and the method comprises the steps: obtaining a first set which comprises the surface temperature of a motor in an integrated pump station at each moment in a historical time period; dividing the data in the first set based on a binary segmentation algorithm, and calculating mark data of each subset obtained after division; removing abnormal data from the first set based on the mark data, and filling missing values in the first set after the abnormal data is removed to obtain a processed first set; and decomposing the processed first set, performing prediction by adopting an unused prediction method after decomposition, and accumulating all prediction results to obtain a surface temperature prediction value of the motor at a future moment. According to the method, processing is carried out in a data preprocessing stage and a model selection stage, and the accuracy of a prediction result can be improved through the mode.

The invention belongs to the technical field of computers, and particularly relates to a state-driven test method for a stability test. Thestate-driven test method comprises the following steps: acquiring the duration of a specified stability test; according to the duration of the stability test, loading a corresponding test data set from the data center module, and initializing a data context according to the loaded test data set; loading a configuration center model according to the initialized data context; according to the configuration center model, selecting a first test case of a preset condition for testing to obtain test result data; updating test result data to a data context, and setting a current state as a next state specified after the function interface is completed in the configuration model; and according to the specified next state, obtaining a test case corresponding to the specified next state, and testing the shown next test case. Stability test modeling is carried out according to test state behaviors, so that the uniformity and authenticity of different testers on stability test models are unified.

The invention discloses a server complete machine performance comparison test method and device, electronic equipment and a computer readable storage medium. The server complete machine performance comparison test method comprises the steps of performing a pressure test on servers participating in a comparison test by gradually increasing concurrent numbers; respectively determining concurrency number limit values when each server enters a performance bottleneck period according to a pressure test result; determining the maximum value in the concurrency number limit values as a target concurrency number; and respectively carrying out pressure test on each server by taking the target concurrency number as a set condition so as to obtain comparison test data. According to the method, the concurrency limit value of each server is obtained through the test case, the comparison test is carried out by taking the maximum value of the concurrency limit value of each server as the test condition, the test data capable of truly and accurately reflecting the performance difference of the whole server is obtained, and a reliable and accurate data basis is provided for the model selection of the server.

The invention discloses an online translation model selection method for statistical machine translation, which includes two stages of training and translation. The training stage includes: collecting bilingual parallel corpora, and dividing the bilingual parallel corpus into different sub-corpora according to types; The corpus trains the candidate translation model; builds an index for the sub-corpus, and obtains the corpus index file; the translation stage includes: inputting the text to be translated, and retrieving sentences similar to the sentences in the text to be translated from the corpus index file; The candidate translation model corresponding to the sub-corpus where the sentence is located, selects the final translation model from all the candidate translation models; translates the input text to be translated according to the final translation model, and obtains the final translation result. The invention well solves the problem that the statistical machine translation system cannot adapt to input texts in different fields, and can effectively improve the translation quality of the statistical machine translation system.

The invention discloses a switching power supply control system of multiple voltage levels, and relates to electronic technology. The system comprises a first level voltage adjusting module, a second level voltage adjusting module, a low-voltage control module and a high-voltage control module, wherein the first level voltage adjusting module comprises a feedback circuit consisting of a depletionmetal oxide semiconductor field effect transistor (MOSFET) device and an operational amplifier, and the source of the depletion MOSFET device is the output end of the first level voltage adjusting module; and a level conversion circuit is connected in series between the low-voltage control module and the high-voltage control module, and the level conversion circuit is also connected with the output end of the first level voltage adjusting module and the output end of the second level voltage adjusting module. The system can provide higher driving voltage at the same time of ensuring the control precision and broaden the model selection range of an external power device.