Cost-sensitive dual-bidirectional linear discriminant analysis

Extracellular Vesicle Preparation and Analysis: A State-of-the-Art Review

In recent decades, research on Extracellular Vesicles (EVs) has gained prominence in the life sciences due to their critical roles in both health and disease states, offering promising applications in disease diagnosis, drug delivery, and therapy. However, their inherent heterogeneity and complex origins pose significant challenges to their preparation, analysis, and subsequent clinical application. This review is structured to provide an overview of the biogenesis, composition, and various sources of EVs, thereby laying the groundwork for a detailed discussion of contemporary techniques for their preparation and analysis. Particular focus is given to state-of-the-art technologies that employ both microfluidic and non-microfluidic platforms for EV processing. Furthermore, this discourse extends into innovative approaches that incorporate artificial intelligence and cutting-edge electrochemical sensors, with a particular emphasis on single EV analysis. This review proposes current challenges and outlines prospective avenues for future research. The objective is to motivate researchers to innovate and expand methods for the preparation and analysis of EVs, fully unlocking their biomedical potential.

Advances of machine learning-assisted small extracellular vesicles detection strategy

Detection of extracellular vesicles (EVs), particularly small EVs (sEVs), is of great significance in exploring their physiological characteristics and clinical applications. The heterogeneity of sEVs plays a crucial role in distinguishing different types of cells and diseases. Machine learning, with its exceptional data processing capabilities, offers a solution to overcome the limitations of conventional detection methods for accurately classifying sEV subtypes and sources. Principal component analysis, linear discriminant analysis, partial least squares discriminant analysis, XGBoost, support vector machine, k-nearest neighbor, and deep learning, along with some combined methods such as principal component-linear discriminant analysis, have been successfully applied in the detection and identification of sEVs. This review focuses on machine learning-assisted detection strategies for cell identification and disease prediction via sEVs, and summarizes the integration of these strategies with surface-enhanced Raman scattering, electrochemistry, inductively coupled plasma mass spectrometry and fluorescence. The performance of different machine learning-based detection strategies is compared, and the advantages and limitations of various machine learning models are also evaluated. Finally, we discuss the merits and limitations of the current approaches and briefly outline the perspective of potential research directions in the field of sEV analysis based on machine learning.

Plasmonic Probing of Deoxyribonucleic Acid Hybridization at the Single Base Pair Resolution

Partial DNA duplex formation greatly impacts the quality of DNA hybridization and has been extensively studied due to its significance in many biological processes. However, traditional DNA sensing methods suffer from time-consuming amplification steps and hinder the acquisition of information about single-molecule behavior. In this work, we developed a plasmonic method to probe the hybridization process at a single base pair resolution and study the relationship between the complementarity of DNA analytes and DNA hybridization behaviors. We measured single-molecule hybridization events with Au NP-modified ssDNA probes in real time and found two hybridization adsorption events: stable and transient adsorption. The ratio of these two hybridization adsorption events was correlated with the length of the complementary sequences, distinguishing DNA analytes from different complementary sequences. By using dual incident angle excitation, we recognized different single-base complementary sequences. These results demonstrated that the plasmonic method can be applied to study partial DNA hybridization behavior and has the potential to be incorporated into the identification of similar DNA sequences, providing a sensitive and quantitative tool for DNA analysis.

Ratio Sum Versus Sum Ratio for Linear Discriminant Analysis

Dimension reduction is a critical technology for high-dimensional data processing, where Linear Discriminant Analysis (LDA) and its variants are effective supervised methods. However, LDA prefers to feature with smaller variance, which causes feature with weak discriminative ability retained. In this paper, we propose a novel Ratio Sum for Linear Discriminant Analysis (RSLDA), which aims at maximizing discriminative ability of each feature in subspace. To be specific, it maximizes the sum of ratio of the between-class distance to the within-class distance in each dimension of subspace. Since the original RSLDA problem is difficult to obtain the closed solution, an equivalent problem is developed which can be solved by an alternative optimization algorithm. For solving the equivalent problem, it is transformed into two sub-problems, one of which can be solved directly, the other is changed into a convex optimization problem, where singular value decomposition is employed instead of matrix inversion. Consequently, performance of algorithm cannot be affected by the non-singularity of covariance matrix. Furthermore, Kernel RSLDA (KRSLDA) is presented to improve the robustness of RSLDA. Additionally, time complexity of RSLDA and KRSLDA are analyzed. Extensive experiments show that RSLDA and KRSLDA outperforms other comparison methods on toy datasets and multiple public datasets.

A rough set based algorithm for updating the modes in categorical clustering

The categorical clustering problem has attracted much attention especially in the last decades since many real world applications produce categorical data. The k-mode algorithm, proposed since 1998, and its multiple variants were widely used in this context. However, they suffer from a great limitation related to the update of the modes in each iteration. The mode in the last step of these algorithms is randomly selected although it is possible to identify many candidate ones. In this paper, a rough density mode selection method is proposed to identify the adequate modes among a list of candidate ones in each iteration of the k-modes. The proposed method, called Density Rough k-Modes (DRk-M) was experimented using real world datasets extracted from the UCI Machine Learning Repository, the Global Terrorism Database (GTD) and a set of collected Tweets. The DRk-M was also compared to many states of the art clustering methods and has shown great efficiency.

Machine learning assisted dual-channel carbon quantum dots-based fluorescence sensor array for detection of tetracyclines

The detection and differentiation of tetracyclines (TCs) has received increasing attention due to the severe threat they pose to human health and the ecological balance. A dual-channel fluorescence sensor array based on two carbon quantum dots (CDs) was fabricated to distinguish between four TCs, including tetracycline (TC), oxytetracycline (OTC), doxycycline (DOX), and metacycline (MTC). A distinct fluorescence variation pattern (I/I₀) was produced when CDs interacted with the four TCs. This pattern was analyzed by LDA and SVM. This was the first time that SVM was used for data processing of fluorescence sensor arrays. LDA and SVM showed that the array has the capacity for parallel and accurate determination of TCs at concentrations between 1.0 μM and 150 μM. In addition, the interference experiment using metal ions and antibiotics as possible coexisting interference substances proves that the sensor array has excellent selectivity and anti-interference ability. The array was also used for the accurate detection and identification of TCs in binary mixtures, and furthermore, the four TCs were successfully identified in river water and milk samples. Besides, the sensor array successfully identified the four TCs in 72 unknown samples with a 100% accuracy. The results proved that SVM can achieve the same accurate classification and prediction as LDA, and considering its additional advantages, it can be used as an optional supplementary method for data processing, thereby expanding the data processing field.

Abd. Big Data Classification Efficiency Based on Linear Discriminant Analysis

The proliferation of online platforms recently has led to unprecedented increase in data generation; this has given rise to the concept of big data which characterizes data in terms of volume, velocity, variety, and veracity. One of the common multivariate statistical data analysis tools is linear discriminant analysis (LDA) which relies on the concept of obtaining the separation among groups through LDA. The prediction of the class of a given class of data points can be achieved through classification, a supervised learning technique but prior to a classification process, a classification model must first be built using classification algorithms. Several classification algorithms are available for prediction tasks. LDA is commonly used for the reduction of the dimensionality of datasets. In this article, the use of LDA to improve the classification performance of different classification model was presented.