コンピュータ・サイエンス : ACM computing surveys

In single-cell RNA sequencing (scRNA-seq), doublets form when two cells are encapsulated into one reaction volume. The existence of doublets, which appear to be-but are not-real cells, is a key confounder in scRNA-seq data analysis. Computational methods have been developed to detect doublets in scRNA-seq data; however, the scRNA-seq field lacks a comprehensive benchmarking of these methods, making it difficult for researchers to choose an appropriate method for specific analyses. We conducted a systematic benchmark study of nine cutting-edge computational doublet-detection methods. Our study included 16 real datasets, which contained experimentally annotated doublets, and 112 realistic synthetic datasets. We compared doublet-detection methods regarding detection accuracy under various experimental settings, impacts on downstream analyses, and computational efficiencies. Our results show that existing methods exhibited diverse performance and distinct advantages in different aspects. Overall, the DoubletFinder method has the best detection accuracy, and the cxds method has the highest computational efficiency. A record of this paper's transparent peer review process is included in the Supplemental Information.

https://escholarship.org/content/qt6gr648np/qt6gr648np.pdf?t=qo0n7q

Benchmarking Computational Doublet-Detection Methods for Single-Cell RNA Sequencing Data.

Class imbalance with overlap is a very challenging problem in electronic fraud transaction detection. Fraudsters have racked their brains to make a fraud transaction as similar as a genuine one in order to avoid being found. Therefore, lots of data of fraud transactions overlap with genuine transactions so that it is hard to distinguish them. However, most attention has been focused on class imbalance rather than overlapping issues for machine-learning-based methods of fraud transaction detection. This paper proposes a novel hybrid method to handle the problem of class imbalance with overlap based on a divide-and-conquer idea. Firstly, an anomaly detection model is trained on the minority samples for excluding both a few outliers of minority class and lots of majority samples from the original dataset. Then the remaining samples form an overlapping subset that has a low imbalance ratio and a reduced learning interference from both minority class and majority class than the original dataset. After that, this difficult overlapping subset is dealt with a non-linear classifier in order to distinguish them well. To achieve good properties of the overlapping subset, we propose a novel assessment criterion, Dynamic Weighted Entropy (DWE), to evaluate its quality. It is a specially designed trade-off between the number of excluded outliers of minority class and the ratio of class imbalance of overlapping subset. With the help of DWE, time consumption on searching good hyper-parameters is dramatically declined. Extensive experiments on Kaggle fraud detection dataset and a large real electronic transaction dataset demonstrate that our method significantly outperforms state-of-the-art ones.

A hybrid method with dynamic weighted entropy for handling the problem of class imbalance with overlap in credit card fraud detection

Automatic fault classification in photovoltaic modules using Convolutional Neural Networks

The growing success of Machine Learning (ML) is making significant improvements to predictive models, facilitating their integration in various application fields. Despite its growing success, there are some limitations and disadvantages: the most significant is the lack of interpretability that does not allow users to understand how particular decisions are made. Our study focus on one of the best performing and most used models in the Machine Learning framework, the Random Forest model. It is known as an efficient model of ensemble learning, as it ensures high predictive precision, flexibility, and immediacy; it is recognized as an intuitive and understandable approach to the construction process, but it is also considered a Black Box model due to the large number of deep decision trees produced within it. The aim of this research is twofold. We present a survey about interpretative proposal for Random Forest and then we perform a machine learning experiment providing a comparison between two methodologies, inTrees, and NodeHarvest, that represent the main approaches in the rule extraction framework. The proposed experiment compares methods performance on six real datasets covering different data characteristics: n. of observations, balanced/unbalanced response, the presence of categorical and numerical predictors. This study contributes to picture a review of the methods and tools proposed for ensemble tree interpretation, and identify, in the class of rule extraction approaches, the best proposal.

A comparison among interpretative proposals for Random Forests

One of the most widely used biomimicry algorithms is the Particle Swarm Optimization (PSO). Since its introduction in 1995, it has caught the attention of both researchers and academicians as a way of solving various optimization problems, such as in the fields of engineering and medicine, to computer image processing and mission critical operations. PSO has been widely applied in the field of swarm robotics, however, the trend of creating a new variant PSO for each swarm robotic project is alarming. We investigate the basic properties of PSO algorithms relevant to the implementation of swarm robotics and characterize the limitations that promote this trend to manifest. Experiments were conducted to investigate the convergence properties of three PSO variants (original PSO, SPSO and APSO) and the global optimum and local optimal of these PSO algorithms were determined. We were able to validate the existence of premature convergence in these PSO variants by comparing 16 functions implemented alongside the PSO variant. This highlighted the fundamental flaws in most variant PSOs, and signifies the importance of developing a more generalized PSO algorithm to support the implementation of swarm robotics. This is critical in curbing the influx of custom PSO and theoretically addresses the fundamental flaws of the existing PSO algorithm.

Understanding the Limitations of Particle Swarm Algorithm for Dynamic Optimization Tasks: A Survey Towards the Singularity of PSO for Swarm Robotic Applications

ABSTRACT Malaria fever is a potentially fatal disease caused by the Plasmodium parasite. Identifying Plasmodium parasites in blood smear images can help diagnose malaria fever rapidly and precisely. According to the World Health Organization (WHO), there were 241 million malaria cases and 627 000 deaths worldwide in 2020, while 95% of malaria cases and 96% of malaria deaths occurred in Africa. Also in Africa, children that are less than five years old accounted for an estimated 80% of all malaria deaths. To address the menace of malaria, this paper proposes a novel deep learning model, called a data augmentation convolutional neural network (DACNN), trained by reinforcement learning to tackle this problem. The performance of the proposed DACNN model is compared with CNN and directed acyclic graph convolutional neural network (DAGCNN) models. Results show that DACNN outperforms previous studies in processing and classification images. It achieved 94.79% classification accuracy in malaria blood sample images of balanced class dataset obtained from the Kaggle dataset. The proposed model can serve as an effective tool for the detection of malaria parasites in blood smear images.

https://doi.org/10.1080/08839514.2022.2033473

Dada Emmanuel Gbenga

Papers

Understanding the Limitations of Particle Swarm Algorithm for Dynamic Optimization Tasks: A Survey Towards the Singularity of PSO for Swarm Robotic Applications

A Novel Data Augmentation Convolutional Neural Network for Detecting Malaria Parasite in Blood Smear Images