No abstract

This paper analyzes the efficiency of the input and output of R&D activity and the status of its development and management in universities in each region of China and proposes suggestions for improvement. The DEA–Malmquist model was used to analyze the static and dynamic science and technology statistics for universities in each region during 2006–2019 to reveal changes in the input and output of R&D activity. The overall efficiency of the R&D activity of universities in all regions of China was low. Among the 27 regions studied, DEA revealed effective efficiency in 20 regions in 2006, accounting for 74.07%, and in 19 regions in 2019, accounting for 70.37%. The Malmquist index was greater than 1 in 17 regions in 2006–2019, with an average value of 1.023 during 2006–2019. The technological progress of R&D activity in universities in each region plays a major role in the improvement of the overall efficiency. Conclusions: The efficiency of the R&D activity of Chinese universities in all regions is low in general, with large disparities between regions. The R&D activity of Chinese universities lacks scientific management. It is necessary to optimize the allocation of research resources, construct an evaluation system for the efficiency of R&D activity, and offer incentives for research to improve the output and promote the transformation of results in Chinese universities.

The convolutional neural network (CNN) is an important network model for processing time series classification tasks. However, there can be a lot of noise in the data itself. The clarity of spatial feature expression of time‐series samples and the freeness and balance of data samples both have a big effect on how well CNNs can handle time series classification (TSC) tasks. This paper proposes a multi‐pose marking learning algorithm (MPML), which optimizes the spatial representation of data samples through multiple graphical representations of time series data and then solves the problems of freeness and heterogeneity of data samples through labeling and preferential learning of poorly trained sample data. The algorithm improves the model's robustness to noise by improving the spatial expression sentiment of the time series and the labeling learning of the feature data, thereby improving the convolutional neural network's performance to perform the TSC task. We show how well the algorithm works by testing it on 46 datasets from the UCR Time Series Classification Archive and comparing the results.