DeepMind | 利用数据驱动深度学习方法应用于天气预报的前景与危机

使用机器学习（ML）模型进行天气预测已经成为一个流行的研究领域。这些模型的前景——无论是与更传统的数值天气预报（NWP）结合使用，还是单独使用——都是为了在大幅降低计算成本的情况下对天气进行更准确的预测。

这个由@DeepMind的工作人员研究科学家Suman Ravuri主讲的AI for Good Discovery讨论了只使用数据驱动（更具体地说是深度学习）方法的前景和危机。强调了最近与气象局合作的一个关于降水预报的项目作为一个案例研究。虽然这项研究发现我们可以创建一个深度学习模型，该模型明显受到气象局气象学家的青睐，并且在客观的性能指标上表现良好，但我们也发现了很多方法，深度学习系统可以在客观的性能指标上表现良好，但却不能提高决策价值。本讲座讨论了这种失败模式出现的一些原因，同时也提倡对纯数据驱动的模型进行更好的验证。虽然讨论的重点是非常短期的预测，但我们认为其中许多教训也适用于使用机器学习进行天气预测的长期预测。

http://mpvideo.qpic.cn/0bc3oaabcaaadmagzmgwtzqva4gdcfyaaeia.f10002.mp4?dis_k=171ad2a5ff11e00dcf6e588aa83fa744&dis_t=1642483615&vid=wxv_2212151302065799175&format_id=10002&support_redirect=0&mmversion=false

The use of machine learning (ML) models for weather prediction has emerged as a popular area of research. The promise of these models — whether in conjunction with more traditional Numerical Weather Prediction (NWP), or on its own — is that they allow for more accurate predictions of the weather at significantly reduced computational cost.This AI for Good Discovery featuring Suman Ravuri, Staff Research Scientist at @DeepMind , discusses both the promise and perils of using a data-driven (and more specifically deep learning) only approach. It highlights a recent project with the Met Office on precipitation nowcasting as a case study. While the study found that we could create a deep learning model that was significantly preferred by Met Office meteorologists and performed well on objective measures of performance, we also discovered many ways in which deep learning systems can perform well on objective measures of performance without improving decision-making value. This talk discusses some reasons why this failure mode occurs, while also advocating for better verification of purely data-driven models. Although the discussion focuses on very-short-term prediction, we believe that many of these lessons are also applicable to longer-term forecasts using machine learning for weather prediction.🎙 Speaker: Suman Ravuri Staff Research Scientist DeepMind🎙 Moderators: Duncan Watson-Parris Postdoctoral Research Associate University of OxfordPhilip Stier Head of Atmospheric, Oceanic and Planetary Physics University of Oxford

What is AI for Good? The AI for Good series is the leading action-oriented, global & inclusive United Nations platform on AI. The Summit is organized all year, always online, in Geneva by the ITU with XPRIZE Foundation in partnership with over 35 sister United Nations agencies, Switzerland and ACM. The goal is to identify practical applications of AI and scale those solutions for global impact.Disclaimer: The views and opinions expressed are those of the panelists and do not reflect the official policy of the ITU.