Framework for kernel regularization with application to protein clustering.
Ontology highlight
ABSTRACT:
SUBMITTER: Lu F
PROVIDER: S-EPMC1187947 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Traditional regression methods typically consider only covariate information and assume that the observations are mutually independent samples. However, samples usually come from individuals connected by a network in many modern applications. We present a risk minimization formulation for learning from both covariates and network structure in the context of graph kernel regularization. The formulation involves a loss function with a penalty term. This penalty can be used not only to encourage similarity between linked nodes but also lead to improvement over traditional regression models. Furthermore, the penalty can be used with many loss-based predictive methods, such as linear regression with squared loss and logistic regression with log-likelihood loss. Simulations to evaluate the performance of this model in the cases of low dimensions and high dimensions show that our proposed approach outperforms all other benchmarks. We verify this for uniform graph, nonuniform graph, balanced-sample, and unbalanced-sample datasets. The approach was applied to predicting the response values on a 'follow' social network of Tencent Weibo users and on two citation networks (Cora and CiteSeer). Each instance verifies that the proposed method combining covariate information and link structure with the graph kernel regularization can improve predictive performance.
| S-EPMC10071950 | biostudies-literature
Project description:Missing data presents a challenge to clustering algorithms, as traditional methods tend to pad incomplete data first before clustering. To combine the two processes of padding and clustering and improve the clustering accuracy, a generalized fuzzy clustering framework is proposed based on optimal completion strategy (OCS) and nearest prototype strategy (NPS) with four improved algorithms developed. Feature weights are introduced to reduce outliers' influence on the cluster centers, and kernel functions are used to solve the linear indistinguishability problem. The proposed algorithms are evaluated regarding correct clustering rate, iteration number, and external evaluation indexes with nine datasets from the UCI (University of California, Irvine) Machine Learning Repository. The results of the experiment indicate that the clustering accuracy of the feature weighted kernel fuzzy C-means algorithm with NPS (NPS-WKFCM) and feature weighted kernel fuzzy C-means algorithm with OCS (OCS-WKFCM) under varying missing rates is superior to that of seven conventional algorithms. Experiments demonstrate that the enhanced algorithm proposed for clustering incomplete data is superior.
| S-EPMC10588703 | biostudies-literature
Project description:Multiple Kernel Learning (MKL) generalizes SVMs to the setting where one simultaneously trains a linear classifier and chooses an optimal combination of given base kernels. Model complexity is typically controlled using various norm regularizations on the base kernel mixing coefficients. Existing methods neither regularize nor exploit potentially useful information pertaining to how kernels in the input set 'interact'; that is, higher order kernel-pair relationships that can be easily obtained via unsupervised (similarity, geodesics), supervised (correlation in errors), or domain knowledge driven mechanisms (which features were used to construct the kernel?). We show that by substituting the norm penalty with an arbitrary quadratic function Q 0, one can impose a desired covariance structure on mixing weights, and use this as an inductive bias when learning the concept. This formulation significantly generalizes the widely used 1- and 2-norm MKL objectives. We explore the model's utility via experiments on a challenging Neuroimaging problem, where the goal is to predict a subject's conversion to Alzheimer's Disease (AD) by exploiting aggregate information from many distinct imaging modalities. Here, our new model outperforms the state of the art (p-values ⪡ 10-3). We briefly discuss ramifications in terms of learning bounds (Rademacher complexity).
| S-EPMC4189130 | biostudies-literature
Project description:Convex clustering is a promising new approach to the classical problem of clustering, combining strong performance in empirical studies with rigorous theoretical foundations. Despite these advantages, convex clustering has not been widely adopted, due to its computationally intensive nature and its lack of compelling visualizations. To address these impediments, we introduce Algorithmic Regularization, an innovative technique for obtaining high-quality estimates of regularization paths using an iterative one-step approximation scheme. We justify our approach with a novel theoretical result, guaranteeing global convergence of the approximate path to the exact solution under easily-checked non-data-dependent assumptions. The application of algorithmic regularization to convex clustering yields the Convex Clustering via Algorithmic Regularization Paths (CARP) algorithm for computing the clustering solution path. On example data sets from genomics and text analysis, CARP delivers over a 100-fold speed-up over existing methods, while attaining a finer approximation grid than standard methods. Furthermore, CARP enables improved visualization of clustering solutions: the fine solution grid returned by CARP can be used to construct a convex clustering-based dendrogram, as well as forming the basis of a dynamic path-wise visualization based on modern web technologies. Our methods are implemented in the open-source R package clustRviz, available at https://github.com/DataSlingers/clustRviz.
| S-EPMC7518335 | biostudies-literature
Project description:In two recent articles a method has been described for calculating the total energy of large molecules. The method is called the kernel energy method (KEM) and requires knowledge of the crystal structure of interest. Calculations are simplified by adopting the approximation that a full molecule could be represented by smaller kernels of atoms. The KEM was illustrated with peptides ranging in size from 4 to 19 amino acid residues, and was found to deliver accurate results. The use of the KEM does not depend upon a particular choice of basis functions and is applicable across quantum computational methods of differing levels of accuracy. These earlier investigations suggested that the KEM could be used to calculate the ab initio quantum mechanical energy of proteins. An application has been made with the protein insulin, composed of 51 aa. Accurate KEM Hartree-Fock energies are obtained for the separate A and B chains of insulin and for their composite structure in the full insulin molecule. A limited basis is used to make possible calculation of the full insulin molecule, which can be used as a standard of accuracy for the KEM calculation. The KEM result obtained is E(KEM) = -21104.7656 a.u. It differs from a full molecule Hartree-Fock result by only 0.000002%. The solvent molecules can be treated effectively as a separate kernel. The KEM result for the fully solvated insulin molecule is E(KEM) = -26275.4127 a.u., differing from the full molecule Hartree-Fock result by as little as 0.000023%.
| S-EPMC1200310 | biostudies-literature
Project description:MotivationDiverse applications-particularly in tumour subtyping-have demonstrated the importance of integrative clustering techniques for combining information from multiple data sources. Cluster Of Clusters Analysis (COCA) is one such approach that has been widely applied in the context of tumour subtyping. However, the properties of COCA have never been systematically explored, and its robustness to the inclusion of noisy datasets is unclear.ResultsWe rigorously benchmark COCA, and present Kernel Learning Integrative Clustering (KLIC) as an alternative strategy. KLIC frames the challenge of combining clustering structures as a multiple kernel learning problem, in which different datasets each provide a weighted contribution to the final clustering. This allows the contribution of noisy datasets to be down-weighted relative to more informative datasets. We compare the performances of KLIC and COCA in a variety of situations through simulation studies. We also present the output of KLIC and COCA in real data applications to cancer subtyping and transcriptional module discovery.Availability and implementationR packages klic and coca are available on the Comprehensive R Archive Network.Supplementary informationSupplementary data are available at Bioinformatics online.
| S-EPMC7750932 | biostudies-literature
Project description:Due to updating the Lagrangian multiplier twice at each iteration, the symmetric alternating direction method of multipliers (S-ADMM) often performs better than other ADMM-type methods. In practical applications, some proximal terms with positive definite proximal matrices are often added to its subproblems, and it is commonly known that large proximal parameter of the proximal term often results in 'too-small-step-size' phenomenon. In this paper, we generalize the proximal matrix from positive definite to indefinite, and propose a new S-ADMM with indefinite proximal regularization (termed IPS-ADMM) for the two-block separable convex programming with linear constraints. Without any additional assumptions, we prove the global convergence of the IPS-ADMM and analyze its worst-case [Formula: see text] convergence rate in an ergodic sense by the iteration complexity. Finally, some numerical results are included to illustrate the efficiency of the IPS-ADMM.
| S-EPMC5522537 | biostudies-other
Project description:For studying cancer and genetic diseases, the issue of identifying high correlation genes from high-dimensional data is an important problem. It is a great challenge to select relevant biomarkers from gene expression data that contains some important correlation structures, and some of the genes can be divided into different groups with a common biological function, chromosomal location or regulation. In this paper, we propose a penalized accelerated failure time model CHR-DE using a non-convex regularization (local search) with differential evolution (global search) in a wrapper-embedded memetic framework. The complex harmonic regularization (CHR) can approximate to the combination [Formula: see text] and ℓq (1 ≤ q < 2) for selecting biomarkers in group. And differential evolution (DE) is utilized to globally optimize the CHR's hyperparameters, which make CHR-DE achieve strong capability of selecting groups of genes in high-dimensional biological data. We also developed an efficient path seeking algorithm to optimize this penalized model. The proposed method is evaluated on synthetic and three gene expression datasets: breast cancer, hepatocellular carcinoma and colorectal cancer. The experimental results demonstrate that CHR-DE is a more effective tool for feature selection and learning prediction.
| S-EPMC6375558 | biostudies-literature
Project description:The Kernel Energy Method (KEM) may be used to calculate quantum mechanical molecular energy by the use of several model chemistries. Simplification is obtained by mathematically breaking a large molecule into smaller parts, called kernels. The full molecule is reassembled from calculations carried out on the kernels. KEM is as yet untested for RNA, and such a test is the purpose here. The basic kernel for RNA is a nucleotide that in general may differ from those of DNA. RNA is a single strand rather than the double helix of DNA. KEM energy has been calculated for a tRNA, whose crystal structure is known, and which contains 2,565 atoms. The energy is calculated to be E = -108,995.1668 (a.u.), in the Hartree-Fock approximation, using a limited basis. Interaction energies are found to be consistent with the hydrogen-bonding scheme previously found. In this paper, the range of biochemical molecules, susceptible of quantum studies by means of the KEM, have been broadened to include RNA.
| S-EPMC1360573 | biostudies-literature
Project description:Fuzzy C-means clustering algorithm is one of the typical clustering algorithms in data mining applications. However, due to the sensitive information in the dataset, there is a risk of user privacy being leaked during the clustering process. The fuzzy C-means clustering of differential privacy protection can protect the user's individual privacy while mining data rules, however, the decline in availability caused by data disturbances is a common problem of these algorithms. Aiming at the problem that the algorithm accuracy is reduced by randomly initializing the membership matrix of fuzzy C-means, in this paper, the maximum distance method is firstly used to determine the initial center point. Then, the gaussian value of the cluster center point is used to calculate the privacy budget allocation ratio. Additionally, Laplace noise is added to complete differential privacy protection. The experimental results demonstrate that the clustering accuracy and effectiveness of the proposed algorithm are higher than baselines under the same privacy protection intensity.
| S-EPMC7987176 | biostudies-literature