Model starting method and device
A model and incremental data technology, applied in the field of machine learning, can solve the problems of ignoring the improvement of model effect, not having model online iteration, and lacking the utilization of a small amount of start-up data, so as to improve accuracy, reduce data volume requirements and Time cost, the effect of solving the cold start problem
Pending Publication Date: 2019-07-05
TENCENT TECH (SHENZHEN) CO LTD
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] Lack of utilization of a small amount of start-up data: ignoring the room for improving the model effect of the original data in the data collection process, that is, not having the ability to iterate online in the initial stage of the model
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
[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.
[0039] It should be noted that the terms "first" and "second" in the description and claims of the present invention and the above drawings are used to distinguish similar objects, but not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances such that the embodiments of the invention described herein can be practiced in sequences other than those illustrate...
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 invention discloses a model starting method and device. The method comprises: original data used for model starting are acquired; obtaining an enhanced data quantity according to the original data; obtaining an enhancement strategy corresponding to the original data according to the quantity of the enhancement data, wherein the enhancement strategy comprises a data enhancement sub-strategy anda feature enhancement sub-strategy; on the basis of the original data, performing enhancement according to the enhancement strategy to obtain incremental data; performing data screening on the incremental data to obtain target incremental data; and combining the target incremental data and the original data to obtain starting data, and starting the model by using the starting data. According to the method, data enhancement processing is carried out in the processes of original data processing and feature data processing, and the processing results are screened, so that the accuracy of the processing results is improved, the problem of cold start is solved, and the data volume requirement and the time cost in the early use period of the model are reduced.
Description
technical field [0001] The present invention relates to the field of machine learning, in particular to a model starting method and device. Background technique [0002] Model startup refers to the process of building a model from scratch to putting it into use without obtaining sufficient data, and finally entering the process of online iterative update. For a multi-object (user) classification / inference problem, the existing cold start technology focuses more on how to obtain additional information to assist the training model in a way different from the normal process of the model. A common method is to introduce additional information and additional logic to assist. For example, when there is a lack of user behavior data in the recommendation system, it will be inferred based on some inherent registration information of the user, such as age and gender; another example is to allow the user to make some additional choices when the user uses it for the first time to obtai...
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 Applications(China)
IPC IPC(8): G06N20/20
Inventor 曹誉仁梁健宾志豪庄灿杰张晨斌黄泽丰倪晓臻胡嘉琪徐森圣陈奋
Owner TENCENT TECH (SHENZHEN) CO LTD