A deep model processing method and device
A model processing and model technology, applied in the field of machine learning, can solve the problems of poor quality of preliminary deep models and long time-consuming optimization iterations, and achieve high-quality results
Active Publication Date: 2017-12-01
SHENZHEN TENCENT COMP SYST CO LTD
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] In view of this, this application provides a deep model processing method and device, which are used to solve the problem of poor quality of the initial deep model obtained by the existing random initialization method and long time-consuming subsequent optimization iterations
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0031] The following will clearly and completely describe the technical solutions in the embodiments of the application with reference to the drawings in the embodiments of the application. Apparently, the described embodiments are only some of the embodiments of the application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.
[0032] see figure 1 , figure 1 It is a flowchart of a deep model processing method disclosed in the embodiment of this application.
[0033] Such as figure 1 As shown, the method includes:
[0034] Step S100, obtaining a shallow model constructed and trained for the target event;
[0035] Here, the shallow model may be a shallow model that has been constructed and trained by others for the target event. This application can directly obtain the existing shallow model.
[0036] ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The present application discloses a deep model processing method and device, which obtains the shallow model constructed and trained for the target event, and uses the input features and input weight values of the shallow model to construct the input layer of the deep model, and further determines The number of hidden layers of the deep model and the number of nodes of each hidden layer initialize the input weight value of each layer node in the deep model so that the output result of the output layer of the deep model is the same as the output result of the shallow model same or close. This application uses the shallow model structure to construct the input layer of the deep model, and initializes the input weight values of the nodes in each layer of the deep model, so that the output results of the deep model are the same or similar to the output results of the shallow model, thus drawing lessons from the shallow model The model is based on the prior knowledge learned from the training data, so that the quality of the initialized preliminary deep model is the same or similar to that of the trained shallow model.
Description
technical field [0001] The present application relates to the technical field of machine learning, and more specifically, to a deep model processing method and device. Background technique [0002] In the field of machine learning, a better model is obtained by building and training a model, which can be used to predict target events. Among them, the models can be divided into two categories, one is the shallow model, and the other is the deep model. The shallow model is directly composed of the input layer and the output layer, while the deep model is composed of the input layer, several hidden layers and the output layer. The deep model is better than the shallow model in terms of model expression ability. [0003] The training process of the deep model is mainly a multiple iterative optimization algorithm, where each iteration process mainly includes: initializing the input weight value of each layer node in the model, and then randomly taking one or a batch of samples, ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F15/18
Inventor 邹永强金涬李毅郭志懋薛伟肖磊
Owner SHENZHEN TENCENT COMP SYST CO LTD