Regression results of crowding out effects. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Article data. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Overidentification test results. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Quantile regression results. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Instrumental variable regression results. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Other robust regression results. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Baseline regression results. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Results of propensity score matching. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Parallel trend test. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Placebo test. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Descriptive statistics of variables. حسب Pu Miao (12886949)

"…Moreover, there is no interaction between talent subsidies and innovation incentives, or between talent subsidies and high-level talent innovation support, but there is an offsetting effect between the other policies. (3) The positive impact of science and technology talent policies exhibits a time lag, generally ranging from 4 to 5 years. …"

Improved GRU prediction unit structure diagram. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Data expansion diagram of PM-GRU model. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Feature fusion structure diagram. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Average accuracy of different models. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Comparison of errors on the coordinate axis. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Running process of improved YOLOv4-Tiny. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Key CHWs’ performance indicators by focus and time period. حسب Nicole Michelén Ströfer (19269154)

Training steps of the improved YOLOv4-Tiny. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"

Flowchart of SPM-GRU model operation. حسب Fuxing He (19555871)

"…Meanwhile, the improved model achieves a processing efficiency of 85 frames/s. In addition, compared with the traditional trajectory prediction model, the constructed model performs the best in table tennis ball trajectory prediction, with errors of 4.5 mm, 25.3 mm, and 35.58 mm. …"