Wyniki wyszukiwania - Biblioteka Nauki

Ten serwis zostanie wyłączony 2025-02-11.

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: multiple instance learning

Sortuj według:

Ogranicz wyniki do:

Future directions in Multiple Instance Learning

100%

Chiu D. , Gondra I. , Xu T.

Journal of Theoretical and Applied Computer Science

2013

tom Vol. 7, nr 3

29--39

In Multiple Instance Learning, each training sample consists of a set of unlabelled instances. The set as a whole is labeled positive if at least one instance in the set is positive, or negative otherwise. Given such training samples, the goal is to learn either an explicit description of the common positive instance(s) or a bag classifier that can assign labels to bags. Previous research has focused on this standard definition of the problem where instances in a set are independent. This raises a question: if we remove the independence assumption, can we generalize the goal of finding a description of the common instance(s) to that of finding a description of the common pattern(s) among instances? Similarly, can we generate bag classifiers that discriminate based on common pattern(s) among instances instead of just common instance(s)? This question raises many other related questions that have not been yet fully explored in the context of this problem. In this paper we first present a survey of existing methods that work with the standard definition of the problem and then elaborate on the previous question in the hope that researchers will investigate this exciting research direction.

Multiple-instance learning with pairwise instance similarity

86%

Yuan L. , Liu J. , Tang X.

2014

tom Vol. 24, no. 3

567--577

Multiple-Instance Learning (MIL) has attracted much attention of the machine learning community in recent years and many real-world applications have been successfully formulated as MIL problems. Over the past few years, several Instance Selection-based MIL (ISMIL) algorithms have been presented by using the concept of the embedding space. Although they delivered very promising performance, they often require long computation times for instance selection, leading to a low efficiency of the whole learning process. In this paper, we propose a simple and efficient ISMIL algorithm based on the similarity of pairwise instances within a bag. The basic idea is selecting from every training bag a pair of the most similar instances as instance prototypes and then mapping training bags into the embedding space that is constructed from all the instance prototypes. Thus, the MIL problem can be solved with the standard supervised learning techniques, such as support vector machines. Experiments show that the proposed algorithm is more efficient than its competitors and highly comparable with them in terms of classification accuracy. Moreover, the testing of noise sensitivity demonstrates that our MIL algorithm is very robust to labeling noise.