Deep learning parallel computing architecture method and hyper-parameter automatic configuration optimization thereof

Abstract

The invention discloses a deep learning parallel computing architecture method and hyper-parameter automatic configuration optimization thereof, and particularly relates to the field of deep learning.Firstly, a CNN is used for capturing spatial features of all places, and then an SRU is used for capturing sequential features of spatiotemporal data on the basis of the spatial features and used forregression prediction of the spatiotemporal data and hyper-parameter automatic configuration optimization of the spatiotemporal data; the invention has the beneficial effects that 1, regression prediction of spatio-temporal data is realized based on the SRU, so that parallel acceleration of the model to a certain extent is realized, and the time consumed by training and reasoning is reduced; 2, automatic configuration of the hyper-parameters of the model is realized based on a parallel genetic algorithm, so that manpower, energy and time consumed by hyper-parameter configuration are reduced,the hyper-parameters are more reasonable, and the prediction performance of the model is better.

Description

technical field [0001] The invention relates to a deep learning parallel computing framework method and its hyperparameter automatic configuration optimization, and specifically relates to the field of deep learning. Background technique [0002] With the continuous development of society and the advancement of science and technology, deep learning and the Internet of Things are widely used in various scenarios, deeply affecting people's daily life and bringing great convenience to people's life. The Internet of Things can collect various data through sensor nodes, such as various environmental data and traffic data, and these data are generally spatiotemporal data, and their research and prediction have great practical value. However, the amount of these data is huge, and manual analysis will consume a lot of manpower, material resources and time, and will also produce large errors. Due to its own characteristics, deep learning can process these data more conveniently and ...

AI Technical Summary

Problems solved by technology

Manual adjustment requires the tuner to have in-depth professional knowledge. It takes a lot of manpower and time to search and search a lot of experiments, wasting a lot of time and energy, and it is also difficult to find the direction of optimization. However, these methods lead to unreasonable parameter selection and affect the model. performance

The grid search needs to provide corresponding optional parameters, and these parameters also need to be provided artificially, and artificially provided numbers generally choose special rules as parameters, which will also lead to unreasonable parameters.

The method based on genetic algorithm will lead to slow solution speed and the possibility of immature convergence, resulting in unsatisfactory parameter adjustment

Images