仿

Yaqin Zhang1, Yichong Wang1, Jixin Chen1

  • 1State Key Laboratory of Marine Environmental Science / National Observation and Research Station for the Taiwan Strait Marine Ecosystem (T-SMART) / Fujian Provincial Key Laboratory for Coastal Ecology and Environmental Studies / College of the Environment and Ecology, Xiamen University, Xiamen, 361102, China.

概括

使用混合深度学习模型 (TAB) 与新的扭曲损失函数 (DILATE) 改进了精确的藻类繁殖预测. 这种方法提高了预测准确度,并减轻了常见的时间序列问题,如数据复制.