Knowledge Transfer with Low-Quality Data: A Feature Extraction Issue

Quanz, Brian; Huan, Jun; Mishra, Meenakshi

doi:10.1109/tkde.2012.75

Cited by 54 publications

(28 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first category would modify the parameters of a source learning model to improve its accuracy in a target domain [30,31]. The second one would reduce the difference between the source and target distributions to adapt the classifier to the target domain [32,33]. The last one would automatically select the training samples that could give a better model for the target task [34,35].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Sequential Monte Carlo filter based on multiple strategies for a scene specialization classifier

Maâmatou

Chateau

Gazzah

et al. 2016

J Image Video Proc.

View full text Add to dashboard Cite

Transfer learning approaches have shown interesting results by using knowledge from source domains to learn a specialized classifier/detector for a target domain containing unlabeled data or only a few labeled samples. In this paper, we present a new transductive transfer learning framework based on a sequential Monte Carlo filter to specialize a generic classifier towards a specific scene. The proposed framework utilizes different strategies and approximates iteratively the hidden target distribution as a set of samples in order to learn a specialized classifier. These training samples are selected from both source and target domains according to their weight importance, which indicates that they belong to the target distribution. The resulting classifier is applied to pedestrian and car detection on several challenging traffic scenes. The experiments have demonstrated that our solution improves and outperforms several state of the art's specialization algorithms on public datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

“…Futhermore, some solutions concatenated the source dataset with new samples, which increased the dataset size during iterations [30][31][32][33]. Others were limited only to the use of samples extracted from the target domain [28], which resulted in losing pertinent information of source samples.…”

Section: Related Workmentioning

confidence: 99%

Sequential Monte Carlo filter based on multiple strategies for a scene specialization classifier

Maâmatou

Chateau

Gazzah

et al. 2016

J Image Video Proc.

View full text Add to dashboard Cite

show abstract

“…To achieve this goal, Pan et al [7] proposed a Transfer Component Analysis (TCA) method to minimize the reconstruction error of the input data by reducing the discrepancy between the labeled and unlabeled data. Quanz et al [19] have explored knowledge transfer of sparse feature, which is a more restricted procedure and prone to overfitting. In addition, Wang et al [20] extends NMF to cross-domain senario.…”

Section: Related Workmentioning

confidence: 99%

Transfer Nonnegative Matrix Factorization for Image Representation

Wang

Gurrin

2016

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Nonnegative Matrix Factorization (NMF) has received considerable attention due to its psychological and physiological interpretation of naturally occurring data whose representation may be partsbased in the human brain. However, when labeled and unlabeled images are sampled from different distributions, they may be quantized into different basis vector space and represented in different coding vector space, which may lead to low representation fidelity. In this paper, we investigate how to extend NMF to cross-domain scenario. We accomplish this goal through TNMF -a novel semi-supervised transfer learning approach. Specifically, we aim to minimize the distribution divergence between labeled and unlabeled images, and incorporate this criterion into the objective function of NMF to construct new robust representations. Experiments show that TNMF outperforms state-of-the-art methods on real datasets.

show abstract

“…Qaunz et al, 2012, [16] have explored a feature extraction perspective, starting with the popular sparse coding approach which learns a set of higher order features for the data. They presented a novel method, where they did not use any classifier.…”

Section: Review On Feature Selection and Reductionmentioning

confidence: 99%

Literature Review of Feature Selection for Mining Tasks

ShakilPervez¹,

Farid²

2015

IJCA

View full text Add to dashboard Cite

During past few decades, researchers worked on data preprocessing techniques for the datasets. Data preprocessing techniques are needed, where the data are prepared for mining. The performance of data mining algorithms in most cases depends on dataset quality, since low-quality training data may lead to the construction of overfitting or fragile classifiers. Also, scientists worked on data mining areas in both algorithms section and conceptions practice section. But for better results they always used the combined or embedded or hybrid approaches. Scientists used different classifiers in different ways and also got their smoother results by arranging some modification in the algorithms. In this paper we shall describe all possible areas of attribute selection and reduction techniques. Feature selection algorithms broadly fall into three categories: filter models, wrapper models and hybrid models. Practically, scientists do the tasks in two stages for obtaining accuracy and that is, they firstly select the features and then reduce the dimensionality of feature vectors with classifiers through learning. Some promising approaches are indicated here and particular concentration is dedicated to describe different methods from raw level to experts, so that in future one can get significant instruction for further analysis.

show abstract

Knowledge Transfer with Low-Quality Data: A Feature Extraction Issue

Cited by 54 publications

References 23 publications

Sequential Monte Carlo filter based on multiple strategies for a scene specialization classifier

Sequential Monte Carlo filter based on multiple strategies for a scene specialization classifier

Transfer Nonnegative Matrix Factorization for Image Representation

Literature Review of Feature Selection for Mining Tasks

Contact Info

Product

Resources

About