DeepME: Deep Mixture Experts for Large-scale Image Classification

Although deep learning has demonstrated its outstanding performance on image classification, most well-known deep networks make efforts to optimize both their structures and their node weights for recognizing fewer (e.g., no more than 1000) object classes. Therefore, it is attractive to extend or mixture such well-known deep networks to support large-scale image classification. According to our best knowledge, how to adaptively and effectively fuse multiple CNNs for large-scale image classification is still under-explored. On this basis, a deep mixture algorithm is developed to support large-scale image classification in this paper. First, a soft spectral clustering method is developed to construct a two-layer ontology (group layer and category layer) by assigning large numbers of image categories into a set of groups according to their inter-category semantic correlations, where the semantically-related image categories under the neighbouring group nodes may share similar learning complexities. Then, such two-layer ontology is further used to generate the task groups, in which each task group contains partial image categories with similar learning complexities and one particular base deep network is learned. Finally, a gate network is learned to combine all base deep networks with fewer diverse outputs to generate a mixture network with larger outputs. Our experimental results on ImageNet10K have demonstrated that our proposed deep mixture algorithm can achieve very competitive results (top 1 accuracy: 32.13%) on large-scale image classification tasks.

Identifying cell types from single-cell data based on similarities and dissimilarities between cells

Background With the development of the technology of single-cell sequence, revealing homogeneity and heterogeneity between cells has become a new area of computational systems biology research. However, the clustering of cell types becomes more complex with the mutual penetration between different types of cells and the instability of gene expression. One way of overcoming this problem is to group similar, related single cells together by the means of various clustering analysis methods. Although some methods such as spectral clustering can do well in the identification of cell types, they only consider the similarities between cells and ignore the influence of dissimilarities on clustering results. This methodology may limit the performance of most of the conventional clustering algorithms for the identification of clusters, it needs to develop special methods for high-dimensional sparse categorical data. Results Inspired by the phenomenon that same type cells have similar gene expression patterns, but different types of cells evoke dissimilar gene expression patterns, we improve the existing spectral clustering method for clustering single-cell data that is based on both similarities and dissimilarities between cells. The method first measures the similarity/dissimilarity among cells, then constructs the incidence matrix by fusing similarity matrix with dissimilarity matrix, and, finally, uses the eigenvalues of the incidence matrix to perform dimensionality reduction and employs the K-means algorithm in the low dimensional space to achieve clustering. The proposed improved spectral clustering method is compared with the conventional spectral clustering method in recognizing cell types on several real single-cell RNA-seq datasets. Conclusions In summary, we show that adding intercellular dissimilarity can effectively improve accuracy and achieve robustness and that improved spectral clustering method outperforms the traditional spectral clustering method in grouping cells.

Analisis Clustering Virus MERS-CoV Menggunakan Metode Spectral Clustering Dan Algoritma K-Means

<p><em>The MERS-Cov virus has spread to other countries outside Saudi Arabia. This is because the MERS-CoV virus can mutate rapidly so it is feared that it could threaten public health and even world health. This virus develops and becomes an acute respiratory disease and the mortality rate reaches 30% among 536 cases. One way to classify the MERS-CoV virus is by grouping the DNA sequences of the MERS-CoV virus which have similar characteristics and functions. Spectral clustering is a grouping method that can identify DNA gene expression. This method is also able to partition DNA data with a more complex structure than the partition clustering method. The purpose of this study was to analyze the MERS-CoV virus clustering using the spectral clustering method and the k-means algorithm. This study used a quantitative descriptive literature approach. The results showed that the results of clustering using the spectral clustering method and the k-means algorithm produced three clusters and were more homogeneous than clustering using k-means only.</em></p>

METHOD OF SPECTRAL CLUSTERING OF PAYMENTS AND RAW MATERIALS SUPPLY FOR THE COMPLIANCE AUDIT PLANNING

Context. The analytical procedures used in the audit are currently based on data mining techniques. The work solves the problem of increasing the efficiency and effectiveness of analytical audit procedures by clustering based on spectral decomposition. The object of the research is the process of auditing the compliance of payment and supply sequences for raw materials. Objective. The aim of the work is to increase the effectiveness and efficiency of the audit due to the method of spectral clustering of sequences of payment and supply of raw materials while automating procedures for checking their compliance. Method. The vectors of features are generated for the objects of the sequences of payment and supply of raw materials, which are then used in the proposed method. The created method improves the traditional spectral clustering method by automatically determining the number of clusters based on the explained and sample variance rule; automatic determination of the scale parameter based on local scaling (the rule of K-nearest neighbors is used); resistance to noise and random outliers by replacing the k-means method with a modified PAM method, i.e. replacing centroid clustering with medoid clustering. As in the traditional approach, the data can be sparse, and the clusters can have different shapes and sizes. The characteristics of evaluating the quality of spectral clustering are selected. Results. The proposed spectral clustering method was implemented in the MATLAB package. The results obtained made it possible to study the dependence of the parameter values on the quality of clustering. Conclusions. The experiments carried out have confirmed the efficiency of the proposed method and allow us to recommend it for practical use in solving audit problems. Prospects for further research may lie in the creation of intelligent parallel and distributed computer systems for general and special purposes, which use the proposed method for segmentation, machine learning and pattern recognition tasks.

Higher-Order Spectral Clustering for Geometric Graphs

The present paper is devoted to clustering geometric graphs. While the standard spectral clustering is often not effective for geometric graphs, we present an effective generalization, which we call higher-order spectral clustering. It resembles in concept the classical spectral clustering method but uses for partitioning the eigenvector associated with a higher-order eigenvalue. We establish the weak consistency of this algorithm for a wide class of geometric graphs which we call Soft Geometric Block Model. A small adjustment of the algorithm provides strong consistency. We also show that our method is effective in numerical experiments even for graphs of modest size.

A Parameter-Free Spectral Clustering Approach to Coherent Structure Detection in Geophysical Flows

In Lagrangian dynamics, the detection of coherent clusters can help understand the organization of transport by identifying regions with coherent trajectory patterns. Many clustering algorithms, however, rely on user-input parameters, requiring a priori knowledge about the flow and making the outcome subjective. Building on the conventional spectral clustering method of Hadjighasem et al (2016), a new parameter-free spectral clustering approach is developed that automatically identifies parameters and does not require any user-input choices. A noise-based metric for quantifying the coherence of the resulting coherent clusters is also introduced. The parameter-free spectral clustering is applied to two benchmark analytical flows, the Bickley Jet and the asymmetric Duffing oscillator, and to a realistic, numerically-generated oceanic coastal flow. In the latter case, the identified model-based clusters are tested using observed trajectories of real drifters. In all examples, our approach succeeded in performing the partition of the domain into coherent clusters with minimal inter-cluster similarity and maximum intra-cluster similarity. For the coastal flow, the resulting coherent clusters are qualitatively similar over the same phase of the tide on different days and even different years, whereas coherent clusters for the opposite tidal phase are qualitatively different.

Uncertainty Quantification of Trajectory Clustering Applied to Ocean Ensemble Forecasts

Partitioning ocean flows into regions dynamically distinct from their surroundings based on material transport can assist search-and-rescue planning by reducing the search domain. The spectral clustering method partitions the domain by identifying fluid particle trajectories that are similar. The partitioning validity depends on the accuracy of the ocean forecasting, which is subject to several sources of uncertainty: model initialization, limited knowledge of the physical processes, boundary conditions, and forcing terms. Instead of a single model output, multiple realizations are produced spanning a range of potential outcomes, and trajectory clustering is used to identify robust features and quantify the uncertainty of the ensemble-averaged results. First, ensemble statistics are used to investigate the cluster sensitivity to the spectral clustering method free-parameters and the forecast parameters for the analytic Bickley jet, a geostrophic flow model. Then, we analyze an operational coastal ocean ensemble forecast and compare the clustering results to drifter trajectories south of Martha’s Vineyard. This approach identifies regions of low uncertainty where drifters released within a cluster predominantly remain there throughout the window of analysis. Drifters released in regions of high uncertainty tend to either enter neighboring clusters or deviate from all predicted outcomes.