دانلود مقاله ISI انگلیسی شماره 105809
ترجمه فارسی عنوان مقاله

انتخاب محدودیت در یادگیری متریک

عنوان انگلیسی
Constraint selection in metric learning
کد مقاله سال انتشار تعداد صفحات مقاله انگلیسی
105809 2018 13 صفحه PDF
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Knowledge-Based Systems, Volume 146, 15 April 2018, Pages 91-103

ترجمه کلمات کلیدی
یادگیری فعال، انتخاب محدودیت پویا، یادگیری متریک، مقیاس نمونه، یادگیری تصادفی،
کلمات کلیدی انگلیسی
Active learning; Dynamic constraint selection; Metric learning; Sample weighting; Stochastic learning;
پیش نمایش مقاله
پیش نمایش مقاله  انتخاب محدودیت در یادگیری متریک

چکیده انگلیسی

A number of machine learning and knowledge-based algorithms are using a metric, or a distance, in order to compare individuals. The Euclidean distance is usually employed, but it may be more efficient to learn a parametric distance such as Mahalanobis metric. Learning such a metric is a hot topic since more than ten years now, and a number of methods have been proposed to efficiently learn it. However, the nature of the problem makes it quite difficult for large scale data, as well as data for which classes overlap. This paper presents a simple way of improving accuracy and scalability of any iterative metric learning algorithm, where constraints are obtained prior to the algorithm. The proposed approach relies on a loss-dependent weighted selection of constraints that are used for learning the metric. Using the corresponding dedicated loss function, the method clearly allows to obtain better results than state-of-the-art methods, both in terms of accuracy and time complexity. Some experimental results on real world, and potentially large, datasets are demonstrating the effectiveness of our proposition.