几何平均度量学习
Geometric Mean Metric Learning
摘要:
我们重新审视从数据中学习欧几里得度量的任务。我们从第一性原理着手解决这个问题,并将其表述为一个令人惊讶的简单优化问题。事实上,我们的公式甚至承认一个封闭形式的解决方案。这个解决方案具有几个非常有吸引力的性质:(i)通过正定矩阵的黎曼几何的先天几何上诉;(ii)易于解释;(iii)计算速度比广泛使用的LMNN和ITML方法快几个数量级。此外,在标准的基准数据集上,我们的封闭式解决方案始终能够获得更高的分类精度。
英文摘要:
We revisit the task of learning a Euclidean metric from data. We approach this problem from first principles and formulate it as a surprisingly simple optimization problem. Indeed, our formulation even admits a closed form solution. This solution possesses several very attractive properties: (i) an innate geometric appeal through the Riemannian geometry of positive definite matrices; (ii) ease of interpretability; and (iii) computational speed several orders of magnitude faster than the widely used LMNN and ITML methods. Furthermore, on standard benchmark datasets, our closed-form solution consistently attains higher classification accuracy.
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。