| 摘 要: 为实现对VR(虚拟现实)大数据的精准测评，保证测评结果更贴合实际，提出深度强化学习VR大数据的智能测评方法。首先，集中筛查VR大数据智能测评指标，使用Bartlett球形工具，检验指标的可靠度；然后，引进深度强化学习，对测评中的特征性因子进行智能拟合，保证测评结果的一致性；最后，通过回归模型设定一个测评标准值与回归系数，按照回归计算中的一致性计算公式，对测评结果进行评估，实现VR大数据智能测评结果校正。实验结果表明，所提的测评方法在实际应用中，能够实现对不同VR大数据测评的高度一致性，并且得出的测评结果与实际更相符。 |
| 关键词: 深度强化学习；测评方法；智能；大数据；测评因子；VR |
|
中图分类号: TP391
文献标识码: A
|
|
| Intelligent Measurement Method for VR Big Data based on Deep Reinforcement Learning |
|
LIU Baiming
|
(Beijing Potential Big Data Research Institute (PRI ), Beijing 100095, China)
Lait_liu@pri.ac.cn
|
| Abstract: To achieve accurate measurement of VR (Virtual Reality) big data and ensure that the measurement results are practical, this paper proposes an intelligent measurement method for VR big data based on deep reinforcement learning algorithm. First of all, the research focuses on screening the intelligent measurement indicators of VR big data, the reliability of which is then tested with Bartlett spherical tool. Secondly, deep reinforcement learning is introduced to intelligently fit the feature factors in the measurement to ensure the consistency of the measurement results. Finally, a measurement standard value and regression coefficient are set through the regression model, and the measurement results are evaluated according to the consistency calculation formula in the regression calculation, realizing VR big data intelligent measurement results correction. The experimental results show that the proposed measurement method can achieve high consistency in the evaluation of different VR big data in practical applications, and the measurement results are more consistent with the reality. |
| Keywords: deep reinforcement learning; measurement methods; intelligence; big data; measurement factors; VR |
