arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.13203cs.LG
收藏

深度学习气候模拟器中的特征重要性

Feature Importance in a Deep Learning Climate Emulator

Brookhaven National Laboratory(布鲁克海文国家实验室);Los Alamos National Laboratory(洛斯阿拉莫斯国家实验室)

Wei Xu, Xihaier Luo, Yihui Ren, Ji Hwan Park, Shinjae Yoo, Balasubramanya T. Nadiga

摘要:

我们提出了一项使用一类事后局部解释方法的研究,即,“理解”气候的深度学习(DL)模拟器的特征重要性方法。具体来说,我们考虑一个多输入单输出的仿真器,使用DenseNet编码器-解码器架构,并经过训练,使用前36个月(适当过滤)的SST数据预测海表面温度(SST)在1,6和9个月的提前期的年际变化。首先,采用特征重要性方法进行个体预测,以时空识别在所选地理区域和所选预测提前期对模型预测重要的输入特征。在第二步中,我们还通过考虑训练样本上的重要性热图的聚合来检查广义意义上的特征重要性的行为。我们发现:1)气候模拟器在任何地理位置的预测主要取决于其周围的小邻域; 2)预测提前时间越长,“重要性”延伸得越远; 3)对于提前顺序,“重要性”的时间衰减与地理位置无关。通过烧蚀实验验证了这一结论。从气候动力学的角度来看,这些研究结果表明,在我们考虑的空间和时间尺度上,本地过程的主导作用和远程遥相关的作用可以忽略不计。从网络架构的角度来看,我们发现的输入和输出之间的时空关系表明潜在的模型改进。我们讨论我们的方法,其中一些我们正在考虑在正在进行的工作进一步扩展。

英文摘要:

We present a study using a class of post-hoc local explanation methods i.e., feature importance methods for "understanding" a deep learning (DL) emulator of climate. Specifically, we consider a multiple-input-single-output emulator that uses a DenseNet encoder-decoder architecture and is trained to predict interannual variations of sea surface temperature (SST) at 1, 6, and 9 month lead times using the preceding 36 months of (appropriately filtered) SST data. First, feature importance methods are employed for individual predictions to spatio-temporally identify input features that are important for model prediction at chosen geographical regions and chosen prediction lead times. In a second step, we also examine the behavior of feature importance in a generalized sense by considering an aggregation of the importance heatmaps over training samples. We find that: 1) the climate emulator's prediction at any geographical location depends dominantly on a small neighborhood around it; 2) the longer the prediction lead time, the further back the "importance" extends; and 3) to leading order, the temporal decay of "importance" is independent of geographical location. An ablation experiment is adopted to verify the findings. From the perspective of climate dynamics, these findings suggest a dominant role for local processes and a negligible role for remote teleconnections at the spatial and temporal scales we consider. From the perspective of network architecture, the spatio-temporal relations between the inputs and outputs we find suggest potential model refinements. We discuss further extensions of our methods, some of which we are considering in ongoing work.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑