进化架构搜索改进 Sentinel-2 湖泊叶绿素 a 预测模型
Evolutionary Architecture Search for Chlorophyll-$a$ Prediction in Lakes using Sentinel-2
有人拿架构搜索重新跑了一遍已发表的湖泊藻华分类模型，参数少了26倍，AUC还从0.790提到0.820，配方和小代码都开源了。
研究者以一项已发表的 Sentinel-2 藻华分类研究为基准，在固定任务、特征和湖泊级训练/测试划分的前提下用正则化进化做 MLP 架构搜索。仅凭内部交叉验证 AUC 选出的网络，将留出集 AUC 从 0.790 提升到 0.820，准确率从 0.733 提升到 0.748。最优模型只有 409 个可训练参数，比最强的人工设计参考模型少 26 倍，体积仅 1.6 kB。搜索收敛出一套一致配方：单层窄层、RMS 归一化、tanh 激活、步进衰减的 RMSprop 加权重平均。该模型小到可作为星载或机载的藻华筛查触发器使用。
Evolutionary Architecture Search for Chlorophyll-$a$ Prediction in Lakes using Sentinel-2
Small tabular datasets with expert-designed spectral features are the norm in operational Earth observation, and the networks applied to them are typically hand-designed. We revisit one such published model -- a Sentinel-2 algal bloom classifier -- and ask what architecture search adds, holding the task, the features and the lake-level train/test split of the original study fixed. Searching an extended multilayer-perceptron space with regularized evolution, and selecting on inner-cross-validation AUC only, we find networks that improve held-out AUC from 0.790 to 0.820 and accuracy from 0.733 to 0.748 while using 409 trainable parameters, 26 times fewer than the strongest hand-designed reference. The search converges on a consistent recipe -- a single narrow layer, RMS normalisation, $\tanh$ activation, step-decayed RMSprop and weight averaging -- that a practitioner would be unlikely to reach by default. At 1.6\,kB the resulting model is small enough to serve as an onboard screening trigger, which is the setting that motivates the work. Code: https://github.com/VU-AIML/automl4eo-bloom-nas.
