<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Sports video courses, as a teaching method and means, are now widely used in the field of physical education and teaching, enabling people to make better use of sports course resources. To study the utilization rate and the effectiveness of the distribution channel of sports course video resources, optimize the application channel for enhancing sports video resources, and improve the exposure rate and the accuracy rate of the sports video courses being retrieved to facilitate users’ query and browsing. This paper investigates the impact of neural network and migration learning models on the application of sports video courses by comparing the experimental results of GRU4Rec and Caser algorithm frameworks and migration learning algorithms with computer processing, analysis, and understanding techniques for sports video content. The results show that too small an embedding vector for sports videos affects the accuracy of the whole embedding vector, and when the relevant behavioral data is small, it is not possible to learn an embedding vector representation with very low dimensionality, when a larger vector d is a better choice. However, the experimental effect becomes poor when d is too large due to the sparse user behavior data and the loss of behavior information caused by too large d. The method proposed in this paper can be better applied in sports video courses.</p>
</abstract>

Sports video courses, as a teaching method and means, are now widely used in the field of physical education and teaching, enabling people to make better use of sports course resources. To study the utilization rate and the effectiveness of the distribution channel of sports course video resources, optimize the application channel for enhancing sports video resources, and improve the exposure rate and the accuracy rate of the sports video courses being retrieved to facilitate users’ query and browsing. This paper investigates the impact of neural network and migration learning models on the application of sports video courses by comparing the experimental results of GRU4Rec and Caser algorithm frameworks and migration learning algorithms with computer processing, analysis, and understanding techniques for sports video content. The results show that too small an embedding vector for sports videos affects the accuracy of the whole embedding vector, and when the relevant behavioral data is small, it is not possible to learn an embedding vector representation with very low dimensionality, when a larger vector d is a better choice. However, the experimental effect becomes poor when d is too large due to the sparse user behavior data and the loss of behavior information caused by too large d. The method proposed in this paper can be better applied in sports video courses.

Neural Networks and Migration Learning Models in Sports Video Course Applications

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

Sports video courses, as a teaching method and means, are now widely used in the field of physical education and teaching, enabling people to make better use...

Neural Networks and Migration Learning Models in Sports Video Course Applications

Pubblicato online: 15 giu 2023

Pagine: -

Ricevuto: 25 giu 2022

Accettato: 17 dic 2022

DOI: https://doi.org/10.2478/amns.2023.1.00398

Parole chiave
Neural network, Transfer learning, Sports video, Curriculum resources

© 2023 Jing Wang et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Neural Networks and Migration Learning Models in Sports Video Course Applications

Pubblicato online: 15 giu 2023

Pagine: -

Ricevuto: 25 giu 2022

Accettato: 17 dic 2022

DOI: https://doi.org/10.2478/amns.2023.1.00398

Parole chiaveNeural network, Transfer learning, Sports video, Curriculum resources

© 2023 Jing Wang et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Parole chiave
Neural network, Transfer learning, Sports video, Curriculum resources