High-dimensional data present significant challenges such as inadequate sample size, abundance of noise, and the curse of dimensionality, which make many traditional classification algorithms inapplicable. To provide valid inference for such data, it requires finding a noise-free low-dimensional representation that preserves both the underlying manifold structure and discriminative information. However, the existing methods often fail to take full consideration of these requirements. In this article, we introduce a robust supervised spline embedding (RS2E) algorithm for high-dimensional classification. The proposed approach is highlighted in four aspects: 1) it preserves the class-aware submanifold structure in the thin plate spline embedding space; 2) it eliminates noise and outliers to recover the clean manifold by exploiting its intrinsic low complexity; 3) it separates the class-aware submanifolds by maximizing the distance between each data point and the marginal data points of other class-aware submanifolds; and 4) it applies the alternating direction method of multipliers with generalized power iteration to solve the objective function. Promising experimental results on the real-world, generative adversarial network (GAN)-generated and artificially corrupted datasets demonstrate that RS2E outperforms other supervised dimensionality reduction algorithms in terms of classification accuracy.

Nonnegative matrix factorization (NMF) has attracted increasing interest for its high interpretability in recent years. It is shown that the NMF is closely related to fuzzy$k$-means clustering, where the basis matrix represents the cluster centroids. However, most of the existing NMF-based clustering algorithms often have their decomposed centroids deviate away from the data manifold, which potentially undermines the clustering results, especially when the datasets lie on complicated geometric structures. In this article, we present a manifold peaks NMF (MPNMF) for data clustering. The proposed approach has the following advantages: 1) it selects a number of MPs to characterize the backbone of the data manifold; 2) it enforces the centroids to lie on the original data manifold, by restricting each centroid to be a conic combination of a small number of nearby MPs; 3) it generalizes the graph smoothness regularization to guide the local graph construction; and 4) it solves a general problem of quadratic regularized nonnegative least squares (NNLSs) with group$l_0$-norm constraint and further develops an efficient optimization algorithm to solve the objective function of the MPNMF. Extensive experiments on both synthetic and real-world datasets demonstrate the effectiveness of the proposed approach.

Summary The classification of gene expression data is significantly important for medical diagnosis. In recent years, compressive sensing emerges as a popular sparse learning method and has been applied in different areas. It is featured with the sparse representation of data with a few atoms in the dictionary. However, the traditional compressive sensing model only focuses on the relationship among different samples but neglects the relationship among different genes. In order to take into account of the both kinds of correlation, we propose a novel bidirectional compressive sensing model for the classification of gene expression data. Under this model, we develop a novel Bi‐ADMM algorithm with three different variants to solve the optimization problem. The promising experimental results on the real‐world gene expression datasets demonstrate both the effectiveness and efficiency of our proposed approach.

To facilitate the computer-aided medical applications, this paper tries to build better intelligent diagnosis systems with the help of swarm intelligence method. As to the clinical data, a built-in graph structure is constructed with training samples being mapped as labeled vertices and test samples being unlabeled vertices. On the basis of the iterative label propagation algorithm, this paper first introduces a confidence-based random walk learning model, where unlabeled vertices that consistently show high probability (above the confidence threshold) in belonging to one class is treated as labeled vertices in the next iteration. Later motivated by the swarm intelligence, this model is further improved by treating the labeled vertices as real ants in nature and the predefined classes as different ant colonies. A novel labeled ant random walk algorithm is introduced by incorporating the history information of random walk in the form of aggregation pheromone. The proposed algorithms are evaluated with a synthetic data as well as some real-life clinical cases in terms of diagnostic accuracy. Experimental results show the potentiality of the proposed algorithms.

To address the issue of evolutionary data classification, we propose an evolving swarm classification model. It treats each class as an ant colony carrying different type of pheromone. The ant colonies send their members to propagate their unique pheromone on the unlabeled instances, so as to label them for member recruitment. Meanwhile, the unlabeled instances are treated as unlabeled ants, which also have their preferences for joining one of those labeled colonies. We call it homing feedback, and integrate it into the pheromone update process. Afterwards, the natural selection process is carried out to keep a balance between the member recruitment and the ant colony size maintenance. Sufficient experiments demonstrate that our algorithm is effective in the real-world evolutionary classification applications.

BACKGROUND: Understanding the localization of proteins in cells is vital to characterizing their functions and possible interactions. As a result, identifying the (sub)cellular compartment within which a protein is located becomes an important problem in protein classification. This classification issue thus involves predicting labels in a dataset with a limited number of labeled data points available. By utilizing a graph representation of protein data, random walk techniques have performed well in sequence classification and functional prediction; however, this method has not yet been applied to protein localization. Accordingly, we propose a novel classifier in the site prediction of proteins based on random walks on a graph. RESULTS: We propose a graph theory model for predicting protein localization using data generated in yeast and gram-negative (Gneg) bacteria. We tested the performance of our classifier on the two datasets, optimizing the model training parameters by varying the laziness values and the number of steps taken during the random walk. Using 10-fold cross-validation, we achieved an accuracy of above 61% for yeast data and about 93% for gram-negative bacteria. CONCLUSIONS: This study presents a new classifier derived from the random walk technique and applies this classifier to investigate the cellular localization of proteins. The prediction accuracy and additional validation demonstrate an improvement over previous methods, such as support vector machine (SVM)-based classifiers.

With the advance of microarray technology, clustering analysis has become a key tool to make sense of the massive amounts of genes expression data. An artificial ants sleeping model (ASM) and an adaptive artificial ants clustering algorithm (A/sup 4/C) are presented to solve the clustering problem in data mining by simulating the behaviors of social ant colonies. In the ASM model, each datum is represented by an agent. The agents' environment is a two-dimensional grid. In A/sup 4/C, the agents can form into high-quality clusters by making simple moves according to little local information from its neighborhood and the parameters are selected and adjusted adaptively. Experimental results on clustering benchmarks show the ASM and A/sup 4/C are simpler, easier to implement, and more efficient than previous methods.

Summary form only given. A parallel algorithm for EDT transform on linear array with reconfigurable pipeline bus system (LARPBS) is presented. For an image with n/spl times/n pixels, the algorithm can complete the EDT transform in O(nlogn/(c(n)logd(n))) time using n.d(n).c(n) processors, where c(n) and d(n) are parameters satisfying 1/spl les/c(n)/spl les/n , and 1

The Euclidean distance transform (EDT) converts a binary image into one where each pixel has a value equal to its distance to the nearest foreground pixel. Two parallel algorithms for EDT on linear array with reconfigurable pipeline bus system (LARPBS) are presented. For an image with n × n pixels, the first algorithm can complete EDT in O[(log n log log n)/(log log log n)] time using n2 processors. The second algorithm can computethe EDT in O(log n log log n) time using n2/(log log n) processors.

A parallel algorithm for Euclidean distance transform (EDT) on linear array with reconfigurable pipeline bus system (LARPBS) is presented. For an image with n/spl times/n pixels, the algorithm can complete EDT transform in O(n log n/c(n) log d(n)) time using n/spl middot/d(n)/spl middot/c(n) processors, where c(n) and d(n) are parameters satisfying 1/spl les/c(n)/spl les/n, and 1