A Selection Metric for semi-supervised learning based on neighborhood construction

Emadi, Mona; Tanha, Jafar; Shiri, Mohammad Ebrahim; Aghdam, Mehdi Hosseinzadeh

doi:10.1016/j.ipm.2020.102444

Cited by 14 publications

(20 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A different example of distance-based selection for semisupervised methods was proposed in [41]. In this paper, the authors use the neighbourhood close to the decision boundary and assign the label of the selected instances using an agreement between classifier and neighbourhood.…”

Section: B Distance-based Methodsmentioning

confidence: 99%

Applying Efficient Selection Techniques of Unlabeled Instances for Wrapper-Based Semi-Supervised Methods

et al. 2022

View full text Add to dashboard Cite

Semi-supervised learning (SSL) is a machine learning approach that integrates supervised and unsupervised learning mechanisms. This integration may be done in different ways and one possibility is to use a wrapper-based strategy. The main aim of a wrapper-based strategy is to use a small number of labelled instances to create a learning model. Then, this created model is used in a labelling process, where some unlabelled instances are labelled, and consequently, these instances are incorporated into the labelled set. One important aspect of a wrapper-based SSL method is the selection of unlabelled instances to be labelled in the labelling process. In other words, an efficient selection process plays an important role in the design of a wrapper-based SSL method since it can lead to an efficient labelling process, and in turn, the creation of efficient learning models. In this paper, we propose the use of three selection methods that can be applied to wrapper-based SSL methods. The main idea is to use two different selection criteria, prediction confidence or classification agreement with a distance metric, to perform an efficient selection of the unlabelled instances. In order to assess the feasibility of the proposed approach, the selection methods are applied in two wellknown wrapper-based SSL methods, which are: Self-training and Co-training. Additionally, an empirical analysis will be conducted in which we compare the standard Self-training and Co-training methods against the proposed versions of these two SSL methods over 35 classification datasets.

show abstract

Section: B Distance-based Methodsmentioning

confidence: 99%

Applying Efficient Selection Techniques of Unlabeled Instances for Wrapper-Based Semi-Supervised Methods

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The labeled data augmentation method, based on a semisupervised learning strategy, involves incorporating pseudolabeled data during training [28], [29], [30], [31], [32]. Craters sharing similar morphology and geology are typically indicative of the same period.…”

Section: A Data Augmentation Based On Semisupervised Strategiesmentioning

confidence: 99%

An Improved Double-Branch Network for Estimation of Crater Ages Based on Semisupervised Learning and Multi-Source Lunar Data

Hong

Zhong

Tong

et al. 2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

While various methods have been developed to estimate the age of impact craters, such as the crater size frequency distribution and morphology methods. Accurately and efficiently estimating the ages of lunar craters using traditional techniques is challenging due to their complex morphology and large number. As a result, the accuracy of age estimation algorithms for meteorite craters based on deep learning is restricted by factors such as a scarcity of age-labeled data and the complex morphology of these craters. To address these issues, this article presents an enhanced double-branch network for estimating crater ages via semisupervised learning and multisource lunar data. The algorithm consists of three steps: semisupervised training data augmentation, adaptive two-branch feature extraction, and a two-stage crater age classification process. The effectiveness of the improved approach was validated through ablation experiments, resulting in an overall accuracy of 83.7% on the test set of meteorite craters. This is 5.2% higher than the accuracy achieved by the previous deep learning method.

show abstract

“…Cada estudio posee un porcentaje de documentos etiquetados y plantean algoritmos afianzados o una extensión de los mismos, acoplados a los métodos propuestos en su estructura para seudo-etiquetar el resto de documentos. Estos algoritmos son iterativos y no iterativos, los más consolidados son EM [18], Kohonen [16], Cross Validation [17] y DPC [19]; mientras que HCSC [8], Newsmap [4], LDA-W2V [14] y MMSL [15] son algoritmos diseñados por los autores con el fin de generar auto-entrenamiento.…”

Section: Modelos Semi-supervisados Para Clasificación De Documentosunclassified

“…Clasificación (E4).-Una vez que el modelo se encuentra entrenado se procede a la clasificación de documentos de prueba para evaluar su desempeño y precisión, en esta etapa se puede hacer uso de clasificadores supervisados o no supervisados ya que se dispone de un conjunto de documentos etiquetados y seudo-etiquetados con un modelo entrenado. Algunos estudios como [8] [15] [19] [16] y [20] hacen uso del clasificador supervisado SVM, otro clasificador supervisado de uso frecuente es NB identificado en [18] [4] y [17], así también en esta etapa se identifica el uso de aprendizaje profundo con algoritmos de redes neuronales, NN y KNN en [21] [22] [23] y [24].…”

Section: Modelos Semi-supervisados Para Clasificación De Documentosunclassified

“…La tendencia que se identifica en este tipo de modelos de auto-entrenamiento es la sencillez de uso de su estructura y la automatización de su procesamiento de etiquetado para que sea un proceso iterativo, esto los convierte mayoritariamente en modelos con esquemas sin mucha carga de procesamiento. En cuanto a su rendimiento, los modelos con mejor precisión [18] en 94 % y [17] en 93.8 % son los que disponen de mayor cantidad de documentos etiquetados para su entrenamiento, 17 % y 70 %, respectivamente; mientras que los de más baja precisión [19] en 50.7 % y [4] en 57 % son aquellos con menor cantidad de etiquetados 10 % y 0 %. El resto de modelos se encuentran en un rango de precisión promedio al grupo con un número de clases que van de 2 a 20 y con una cantidad de etiquetados del 5 % al 10 %.…”

Section: Auto-entrenamientounclassified

See 1 more Smart Citation

Semi-supervised learning models for document classification: A systematic review and meta-analysis

Cevallos-Culqui,

Pons,

Rodriguez

2023

View full text Add to dashboard Cite

The continuous increase of digital documents on the web creates the need to search for information patterns that allow the categorization of organizational documents to generate knowledge in an institution. An Artificial Intelligence technique for this purpose is text classification, it for its application uses labels (previously categorized documents) with supervised (with labels) or unsupervised (without labels) training models. Both traditional models with their advantages and disadvantages have been joined into semi-supervised models that extract the best qualities of each one, however, the labeling process involves resources and time that try to be optimized to improve classification accuracy. An analysis of the different semi-supervised models would show us the advantages of their training and the way how the structure of each of them affects the accuracy of their classification. In the present study, a classification structure of the semi-supervised models in the classification of documents is proposed to analyze their qualities and categorization process, through an SLR (Revision of systematic literature) that extracts performance metrics from the identified studies to perform a meta-analysis through forest plots. To define the search strategy for studies, the PICOC (Population, Intervention, Comparison, Outputs, Context) method has been used, it is supported by the research question defines a search string, which has allowed the collection of 228 research, these are filtered with the PRISMA declaration method and the determination of exclusion criteria, in this way 35 researches are selected for the present study. The analysis of the selected studies identifies a structure for the different semi-supervised learning models, and a scheme of their work process is obtained, it has been used to extract advantages, disadvantages, and performance metrics. Through a meta-analysis with forest diagrams, the classification accuracy performance of the researches in each learning model is evaluated, determining as results that regardless of the characteristics of its process, active learning (0.89) and assembled learning (0.83) present the best performance levels.

show abstract

A Selection Metric for semi-supervised learning based on neighborhood construction

Cited by 14 publications

References 32 publications

Applying Efficient Selection Techniques of Unlabeled Instances for Wrapper-Based Semi-Supervised Methods

Applying Efficient Selection Techniques of Unlabeled Instances for Wrapper-Based Semi-Supervised Methods

An Improved Double-Branch Network for Estimation of Crater Ages Based on Semisupervised Learning and Multi-Source Lunar Data

Semi-supervised learning models for document classification: A systematic review and meta-analysis

Contact Info

Product

Resources

About