We present a method to compute the derivative of a learning task with respect to a dataset. A learning task is a function from a training set to the validation error, which can be represented by a trained deep neural network (DNN). The "dataset derivative" is a linear operator, computed around the trained model, that informs how perturbations of the weight of each training sample affect the validation error, usually computed on a separate validation dataset. Our method, DIVA (Differentiable Validation) hinges on a closed-form differentiable expression of the leave-one-out cross-validation error around a pre-trained DNN. Such expression constitutes the dataset derivative. DIVA could be used for dataset auto-curation, for example removing samples with faulty annotations, augmenting a dataset with additional relevant samples, or rebalancing. More generally, DIVA can be used to optimize the dataset, along with the parameters of the model, as part of the training process without the need for a separate validation dataset, unlike bi-level optimization methods customary in AutoML. To illustrate the flexibility of DIVA, we report experiments on sample auto-curation tasks such as outlier rejection, dataset extension, and automatic aggregation of multi-modal data.
翻译:我们提出了一个计算与数据集有关的学习任务衍生物的方法。学习任务是一个从训练组到验证错误的函数,可由经过训练的深神经网络(DNN)代表。“数据集衍生物”是一个线性操作员,围绕经过训练的模式计算。“数据集衍生物”是一个线性操作员,根据经过训练的模式计算,告知每个训练样品重量的扰动如何影响验证错误,通常在单独的验证数据集中计算。我们的方法,DIVA(不同的校验)取决于在预先训练的DNN(DN)周围的留置一出交叉校验错误的封闭式不同表达方式。这种表达方式构成数据集衍生物。DIVA(DIVA)可以用于数据集自动校验,例如删除带有错误说明的样品,用额外的相关样品补充数据集,或重新平衡。更一般地,DIVA(不同的校验)可以用来优化数据集以及模型参数,作为培训过程的一部分,而无需单独鉴定数据集,与AutalML(AutalML)的双级优化方法不同。为了说明DIVA(DIVA)的自动数据模型的扩展任务,我们报告这种数据模拟数据模拟的自动抽样实验。