Using Deep Learning to Classify Class Imbalanced Gene-Expression Microarrays Datasets

Reyes-Nava, A.; Cruz-Reyes, H.; Alejo, R.; Rendón, Eréndira; Flores-Fuentes, A. A.; Granda‐Gutiérrez, E. E.

doi:10.1007/978-3-030-13469-3_6

Cited by 8 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In machine learning scenarios, SMOTE traditionally overcomes ROS's performance [24], but in the big data and deep learning context, results (of this work) show that this expectation was not sufficiently fulfilled. This behavior has been observed in other works related to big data and deep learning [1,48]. However, the performances of both ROS and SMOTE were practically equivalent.…”

Section: Methodssupporting

confidence: 80%

“…It has seen that big data class imbalance approaches have been addressed by adaptation of traditional techniques, mainly sampling methods [21,44]. However, recent studies show that some conclusions from machine learning are not applicable to the big data context; for example, in machine learning is common that SMOTE performs better than ROS [24], but in big data some results do not show this trend [1,48]. In addition, only a few works have been addressed to deal with the class imbalance in big data by using "intelligent" or heuristic sampling techniques [17,49].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

et al. 2020

View full text Add to dashboard Cite

The class imbalance problem has been a hot topic in the machine learning community in recent years. Nowadays, in the time of big data and deep learning, this problem remains in force. Much work has been performed to deal to the class imbalance problem, the random sampling methods (over and under sampling) being the most widely employed approaches. Moreover, sophisticated sampling methods have been developed, including the Synthetic Minority Over-sampling Technique (SMOTE), and also they have been combined with cleaning techniques such as Editing Nearest Neighbor or Tomek's Links (SMOTE+ENN and SMOTE+TL, respectively). In the big data context, it is noticeable that the class imbalance problem has been addressed by adaptation of traditional techniques, relatively ignoring intelligent approaches. Thus, the capabilities and possibilities of heuristic sampling methods on deep learning neural networks in big data domain are analyzed in this work, and the cleaning strategies are particularly analyzed. This study is developed on big data, multi-class imbalanced datasets obtained from hyper-spectral remote sensing images. The effectiveness of a hybrid approach on these datasets is analyzed, in which the dataset is cleaned by SMOTE followed by the training of an Artificial Neural Network (ANN) with those data, while the neural network output noise is processed with ENN to eliminate output noise; after that, the ANN is trained again with the resultant dataset. Obtained results suggest that best classification outcome is achieved when the cleaning strategies are applied on an ANN output instead of input feature space only. Consequently, the need to consider the classifier's nature when the classical class imbalance approaches are adapted in deep learning and big data scenarios is clear.

show abstract

Section: Methodssupporting

confidence: 80%

Section: Introductionmentioning

confidence: 99%

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

et al. 2020

View full text Add to dashboard Cite

show abstract

“…To illustrate, it is normal for machine learning that SMOTE can provide better results than ROS [10]. However, in some cases, the results did not represent similar trends in big data contexts [11] [12]. Additionally, not so many previous studies have focused on big data class imbalance resolution towards the application of "intelligent" or heuristic sampling methods [13] [14].…”

Section: Literature Reviewmentioning

confidence: 99%

“…Precision TP TP FP (9) Recall TP TP FN (10) F measure Precision Recall Precision Recall u u 2 (11) where: TP refers to when the targeted class is "Yes" and the model predicts it "Yes". TN refers to when the targeted class is "No" and the model predicts it "No".…”

Section: Accuracy Tp Tn Tp Tn Fp Fn (8)mentioning

confidence: 99%

Feature Selection for Analyzing Data Errors Toward Development of Household Big Data at the Sub-District Level Using Multi-Layer Perceptron Neural Network

Nuanmeesri

2022

Int. J. Interact. Mob. Technol.

View full text Add to dashboard Cite

This research aims to analyze the patterns of data errors in order to fulfill the data required for household big data development at the sub-district level in Thailand. Feature Selection and Multi-Layer Perceptron Neural Network were applied, while the data imbalance was solved by the SMOTE method and the comparison between the CFS feature selection method and Information Gain (IG) feature selection method. Afterward, the datasets were classified the data errors by the Multi-Layer Perceptron Neural Network. Each model’s effectiveness was measured by the 10-fold cross-validation method. The research results revealed that the suitable data size after being adjusted data imbalanced was 400%. Once the data had been processed for developing the model, it was found that after being adjusted data size towards the application of the SMOTE, CFS feature selection technique, and classified data errors by the Multi-Layer Perceptron Neural Network, the model provided the highest level of effectiveness in data errors classification with an accuracy of 98.29 %. Moreover, the application could effectively classify data errors and display the household big data at the highest level. The application evaluation results given by the experts and the users had an average mean of 4.69 and higher, a standard deviation of 0.47 and lower, which has the level of effectiveness of 93.78% and higher, while interquartile range values not over 1, a quartile deviation of no more than 0.5.

show abstract

“…Considering the specific focus of this paper, we shall cover only relevant articles that use feature selection approaches to tackle high-dimensional multi-class imbalanced datasets. Lately, feature reduction techniques such as feature selection have been applied to tackle the imbalanced class problem at the feature level [6] because most of the high dimensional data sets have imbalanced class problems [6,74], such as bioinformatics [75], text categorization [76], microarray data set [77].…”

Section: Related Workmentioning

confidence: 99%

Multiple Filter-Based Rankers to Guide Hybrid Grasshopper Optimization Algorithm and Simulated Annealing for Feature Selection With High Dimensional Multi-Class Imbalanced Datasets

Sharifai

Zainol

2021

IEEE Access

View full text Add to dashboard Cite

DNA microarray data analysis is infamous due to a massive number of features, imbalanced class distribution, and limited available samples. In this paper, we focus on high-dimensional multi-class imbalanced problems. The high dimensional and multi-class imbalanced problem has posed acute challenges for the conventional classifiers to effectively perform classification tasks on both the minority and majority classes. Numerous efforts have been devoted to addressing either high dimensionality dataset or class imbalance problems. Nonetheless, few methods have been proposed to address the intersection of multi-class imbalanced and high-dimensional problems concurrently due to their intricate interactions. This paper presents novel hybrid algorithms for feature selection with the high dimensional multi-class imbalanced problem using multiple filter-based rankers (MFR) and hybrid Grasshopper optimization algorithm (GOA). The Simulated Annealing (SA) algorithm is incorporated into GOA. SA is used to enhance the best solution found by the GOA algorithm. The aim of using the SA here is to tackle the slow convergence and improve the exploitation by searching the high-quality regions found by the GOA. The experimental results confirm the effectiveness of the proposed methods in improving the classification performance in terms of area under the curve (AUC) compared to other well-known methods, which guarantees the ability of the proposed methods in searching the feature space and identifying very robust and discriminative features that best predict the minority class.

show abstract

Using Deep Learning to Classify Class Imbalanced Gene-Expression Microarrays Datasets

Cited by 8 publications

References 20 publications

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

Feature Selection for Analyzing Data Errors Toward Development of Household Big Data at the Sub-District Level Using Multi-Layer Perceptron Neural Network

Multiple Filter-Based Rankers to Guide Hybrid Grasshopper Optimization Algorithm and Simulated Annealing for Feature Selection With High Dimensional Multi-Class Imbalanced Datasets

Contact Info

Product

Resources

About