Armin Askari, Quentin Rebjock, A. d’Aspremont, L. Ghaoui
We describe a series of algorithms that efficiently implement Gaussian model-X knockoffs to control the false discovery rate on large scale feature selection problems. Identifying the knockoff distribution requires solving a large scale semidefinite program for which we derive several efficient methods. One handles generic covariance matrices, has a complexity scaling as $O(p^3)$ where $p$ is the ambient dimension, while another assumes a rank $k$ factor model on the covariance matrix to reduce this complexity bound to $O(pk^2)$. We also derive efficient procedures to both estimate factor models and sample knockoff covariates with complexity linear in the dimension. We test our methods on problems with $p$ as large as $500,000$.
{"title":"FANOK: Knockoffs in Linear Time","authors":"Armin Askari, Quentin Rebjock, A. d’Aspremont, L. Ghaoui","doi":"10.1137/20m1363698","DOIUrl":"https://doi.org/10.1137/20m1363698","url":null,"abstract":"We describe a series of algorithms that efficiently implement Gaussian model-X knockoffs to control the false discovery rate on large scale feature selection problems. Identifying the knockoff distribution requires solving a large scale semidefinite program for which we derive several efficient methods. One handles generic covariance matrices, has a complexity scaling as $O(p^3)$ where $p$ is the ambient dimension, while another assumes a rank $k$ factor model on the covariance matrix to reduce this complexity bound to $O(pk^2)$. We also derive efficient procedures to both estimate factor models and sample knockoff covariates with complexity linear in the dimension. We test our methods on problems with $p$ as large as $500,000$.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"22 1","pages":"833-853"},"PeriodicalIF":0.0,"publicationDate":"2020-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86856742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Motivated by surprisingly good generalization properties of learned deep neural networks in overparameterized scenarios and by the related double descent phenomenon, this paper analyzes the relation between smoothness and low generalization error in an overparameterized linear learning problem. We study a random Fourier series model, where the task is to estimate the unknown Fourier coefficients from equidistant samples. We derive exact expressions for the generalization error of both plain and weighted least squares estimators. We show precisely how a bias towards smooth interpolants, in the form of weighted trigonometric interpolation, can lead to smaller generalization error in the overparameterized regime compared to the underparameterized regime. This provides insight into the power of overparameterization, which is common in modern machine learning.
{"title":"Overparameterization and Generalization Error: Weighted Trigonometric Interpolation","authors":"Yuege Xie, H. Chou, H. Rauhut, Rachel A. Ward","doi":"10.1137/21m1390955","DOIUrl":"https://doi.org/10.1137/21m1390955","url":null,"abstract":"Motivated by surprisingly good generalization properties of learned deep neural networks in overparameterized scenarios and by the related double descent phenomenon, this paper analyzes the relation between smoothness and low generalization error in an overparameterized linear learning problem. We study a random Fourier series model, where the task is to estimate the unknown Fourier coefficients from equidistant samples. We derive exact expressions for the generalization error of both plain and weighted least squares estimators. We show precisely how a bias towards smooth interpolants, in the form of weighted trigonometric interpolation, can lead to smaller generalization error in the overparameterized regime compared to the underparameterized regime. This provides insight into the power of overparameterization, which is common in modern machine learning.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47917183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We present a new approach to solve the sparse approximation or best subset selection problem, namely find a $k$-sparse vector ${bf x}inmathbb{R}^d$ that minimizes the $ell_2$ residual $lVert A{bf x}-{bf y} rVert_2$. We consider a regularized approach, whereby this residual is penalized by the non-convex $textit{trimmed lasso}$, defined as the $ell_1$-norm of ${bf x}$ excluding its $k$ largest-magnitude entries. We prove that the trimmed lasso has several appealing theoretical properties, and in particular derive sparse recovery guarantees assuming successful optimization of the penalized objective. Next, we show empirically that directly optimizing this objective can be quite challenging. Instead, we propose a surrogate for the trimmed lasso, called the $textit{generalized soft-min}$. This penalty smoothly interpolates between the classical lasso and the trimmed lasso, while taking into account all possible $k$-sparse patterns. The generalized soft-min penalty involves summation over $binom{d}{k}$ terms, yet we derive a polynomial-time algorithm to compute it. This, in turn, yields a practical method for the original sparse approximation problem. Via simulations, we demonstrate its competitive performance compared to current state of the art.
{"title":"The Trimmed Lasso: Sparse Recovery Guarantees and Practical Optimization by the Generalized Soft-Min Penalty","authors":"Tal Amir, R. Basri, B. Nadler","doi":"10.1137/20M1330634","DOIUrl":"https://doi.org/10.1137/20M1330634","url":null,"abstract":"We present a new approach to solve the sparse approximation or best subset selection problem, namely find a $k$-sparse vector ${bf x}inmathbb{R}^d$ that minimizes the $ell_2$ residual $lVert A{bf x}-{bf y} rVert_2$. We consider a regularized approach, whereby this residual is penalized by the non-convex $textit{trimmed lasso}$, defined as the $ell_1$-norm of ${bf x}$ excluding its $k$ largest-magnitude entries. We prove that the trimmed lasso has several appealing theoretical properties, and in particular derive sparse recovery guarantees assuming successful optimization of the penalized objective. Next, we show empirically that directly optimizing this objective can be quite challenging. Instead, we propose a surrogate for the trimmed lasso, called the $textit{generalized soft-min}$. This penalty smoothly interpolates between the classical lasso and the trimmed lasso, while taking into account all possible $k$-sparse patterns. The generalized soft-min penalty involves summation over $binom{d}{k}$ terms, yet we derive a polynomial-time algorithm to compute it. This, in turn, yields a practical method for the original sparse approximation problem. Via simulations, we demonstrate its competitive performance compared to current state of the art.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"47 10 1","pages":"900-929"},"PeriodicalIF":0.0,"publicationDate":"2020-05-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80619433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Or Yair, Felix Dietrich, Rotem Mulayoff, R. Talmon, I. Kevrekidis
In this paper, we propose a spectral method for deriving functions that are jointly smooth on multiple observed manifolds. Our method is unsupervised and primarily consists of two steps. First, using kernels, we obtain a subspace spanning smooth functions on each manifold. Then, we apply a spectral method to the obtained subspaces and discover functions that are jointly smooth on all manifolds. We show analytically that our method is guaranteed to provide a set of orthogonal functions that are as jointly smooth as possible, ordered from the smoothest to the least smooth. In addition, we show that the proposed method can be efficiently extended to unseen data using the Nystrom method. We demonstrate the proposed method on both simulated and real measured data and compare the results to nonlinear variants of the seminal Canonical Correlation Analysis (CCA). Particularly, we show superior results for sleep stage identification. In addition, we show how the proposed method can be leveraged for finding minimal realizations of parameter spaces of nonlinear dynamical systems.
{"title":"Spectral Discovery of Jointly Smooth Features for Multimodal Data","authors":"Or Yair, Felix Dietrich, Rotem Mulayoff, R. Talmon, I. Kevrekidis","doi":"10.1137/21M141590X","DOIUrl":"https://doi.org/10.1137/21M141590X","url":null,"abstract":"In this paper, we propose a spectral method for deriving functions that are jointly smooth on multiple observed manifolds. Our method is unsupervised and primarily consists of two steps. First, using kernels, we obtain a subspace spanning smooth functions on each manifold. Then, we apply a spectral method to the obtained subspaces and discover functions that are jointly smooth on all manifolds. We show analytically that our method is guaranteed to provide a set of orthogonal functions that are as jointly smooth as possible, ordered from the smoothest to the least smooth. In addition, we show that the proposed method can be efficiently extended to unseen data using the Nystrom method. We demonstrate the proposed method on both simulated and real measured data and compare the results to nonlinear variants of the seminal Canonical Correlation Analysis (CCA). Particularly, we show superior results for sleep stage identification. In addition, we show how the proposed method can be leveraged for finding minimal realizations of parameter spaces of nonlinear dynamical systems.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"1 1","pages":"410-430"},"PeriodicalIF":0.0,"publicationDate":"2020-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90987478","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
K. Khamaru, A. Pananjady, Feng Ruan, M. Wainwright, Michael I. Jordan
We address the problem of policy evaluation in discounted Markov decision processes, and provide instance-dependent guarantees on the $ell_infty$-error under a generative model. We establish both asymptotic and non-asymptotic versions of local minimax lower bounds for policy evaluation, thereby providing an instance-dependent baseline by which to compare algorithms. Theory-inspired simulations show that the widely-used temporal difference (TD) algorithm is strictly suboptimal when evaluated in a non-asymptotic setting, even when combined with Polyak-Ruppert iterate averaging. We remedy this issue by introducing and analyzing variance-reduced forms of stochastic approximation, showing that they achieve non-asymptotic, instance-dependent optimality up to logarithmic factors.
{"title":"Is Temporal Difference Learning Optimal? An Instance-Dependent Analysis","authors":"K. Khamaru, A. Pananjady, Feng Ruan, M. Wainwright, Michael I. Jordan","doi":"10.1137/20m1331524","DOIUrl":"https://doi.org/10.1137/20m1331524","url":null,"abstract":"We address the problem of policy evaluation in discounted Markov decision processes, and provide instance-dependent guarantees on the $ell_infty$-error under a generative model. We establish both asymptotic and non-asymptotic versions of local minimax lower bounds for policy evaluation, thereby providing an instance-dependent baseline by which to compare algorithms. Theory-inspired simulations show that the widely-used temporal difference (TD) algorithm is strictly suboptimal when evaluated in a non-asymptotic setting, even when combined with Polyak-Ruppert iterate averaging. We remedy this issue by introducing and analyzing variance-reduced forms of stochastic approximation, showing that they achieve non-asymptotic, instance-dependent optimality up to logarithmic factors.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"1 1","pages":"1013-1040"},"PeriodicalIF":0.0,"publicationDate":"2020-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73570067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
L. Cowen, K. Devkota, Xiaozhe Hu, James M. Murphy, Kaiyi Wu
Data-dependent metrics are powerful tools for learning the underlying structure of high-dimensional data. This article develops and analyzes a data-dependent metric known as diffusion state distance (DSD), which compares points using a data-driven diffusion process. Unlike related diffusion methods, DSDs incorporate information across time scales, which allows for the intrinsic data structure to be inferred in a parameter-free manner. This article develops a theory for DSD based on the multitemporal emergence of mesoscopic equilibria in the underlying diffusion process. New algorithms for denoising and dimension reduction with DSD are also proposed and analyzed. These approaches are based on a weighted spectral decomposition of the underlying diffusion process, and experiments on synthetic datasets and real biological networks illustrate the efficacy of the proposed algorithms in terms of both speed and accuracy. Throughout, comparisons with related methods are made, in order to illustrate the distinct advantages of DSD for datasets exhibiting multiscale structure.
{"title":"Diffusion State Distances: Multitemporal Analysis, Fast Algorithms, and Applications to Biological Networks","authors":"L. Cowen, K. Devkota, Xiaozhe Hu, James M. Murphy, Kaiyi Wu","doi":"10.1137/20M1324089","DOIUrl":"https://doi.org/10.1137/20M1324089","url":null,"abstract":"Data-dependent metrics are powerful tools for learning the underlying structure of high-dimensional data. This article develops and analyzes a data-dependent metric known as diffusion state distance (DSD), which compares points using a data-driven diffusion process. Unlike related diffusion methods, DSDs incorporate information across time scales, which allows for the intrinsic data structure to be inferred in a parameter-free manner. This article develops a theory for DSD based on the multitemporal emergence of mesoscopic equilibria in the underlying diffusion process. New algorithms for denoising and dimension reduction with DSD are also proposed and analyzed. These approaches are based on a weighted spectral decomposition of the underlying diffusion process, and experiments on synthetic datasets and real biological networks illustrate the efficacy of the proposed algorithms in terms of both speed and accuracy. Throughout, comparisons with related methods are made, in order to illustrate the distinct advantages of DSD for datasets exhibiting multiscale structure.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"32 1","pages":"142-170"},"PeriodicalIF":0.0,"publicationDate":"2020-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80648699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kernel methods have achieved very good performance on large scale regression and classification problems, by using the Nystrom method and preconditioning techniques. The Nystrom approximation -- based on a subset of landmarks -- gives a low rank approximation of the kernel matrix, and is known to provide a form of implicit regularization. We further elaborate on the impact of sampling diverse landmarks for constructing the Nystrom approximation in supervised as well as unsupervised kernel methods. By using Determinantal Point Processes for sampling, we obtain additional theoretical results concerning the interplay between diversity and regularization. Empirically, we demonstrate the advantages of training kernel methods based on subsets made of diverse points. In particular, if the dataset has a dense bulk and a sparser tail, we show that Nystrom kernel regression with diverse landmarks increases the accuracy of the regression in sparser regions of the dataset, with respect to a uniform landmark sampling. A greedy heuristic is also proposed to select diverse samples of significant size within large datasets when exact DPP sampling is not practically feasible.
{"title":"Diversity sampling is an implicit regularization for kernel methods","authors":"M. Fanuel, J. Schreurs, J. Suykens","doi":"10.1137/20M1320031","DOIUrl":"https://doi.org/10.1137/20M1320031","url":null,"abstract":"Kernel methods have achieved very good performance on large scale regression and classification problems, by using the Nystrom method and preconditioning techniques. The Nystrom approximation -- based on a subset of landmarks -- gives a low rank approximation of the kernel matrix, and is known to provide a form of implicit regularization. We further elaborate on the impact of sampling diverse landmarks for constructing the Nystrom approximation in supervised as well as unsupervised kernel methods. By using Determinantal Point Processes for sampling, we obtain additional theoretical results concerning the interplay between diversity and regularization. Empirically, we demonstrate the advantages of training kernel methods based on subsets made of diverse points. In particular, if the dataset has a dense bulk and a sparser tail, we show that Nystrom kernel regression with diverse landmarks increases the accuracy of the regression in sparser regions of the dataset, with respect to a uniform landmark sampling. A greedy heuristic is also proposed to select diverse samples of significant size within large datasets when exact DPP sampling is not practically feasible.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"9 1","pages":"280-297"},"PeriodicalIF":0.0,"publicationDate":"2020-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85923398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Samuel Horváth, Lihua Lei, Peter Richtárik, Michael I. Jordan
Adaptivity is an important yet under-studied property in modern optimization theory. The gap between the state-of-the-art theory and the current practice is striking in that algorithms with desirable theoretical guarantees typically involve drastically different settings of hyperparameters, such as step-size schemes and batch sizes, in different regimes. Despite the appealing theoretical results, such divisive strategies provide little, if any, insight to practitioners to select algorithms that work broadly without tweaking the hyperparameters. In this work, blending the "geometrization" technique introduced by Lei & Jordan 2016 and the texttt{SARAH} algorithm of Nguyen et al., 2017, we propose the Geometrized texttt{SARAH} algorithm for non-convex finite-sum and stochastic optimization. Our algorithm is proved to achieve adaptivity to both the magnitude of the target accuracy and the Polyak-Łojasiewicz (PL) constant if present. In addition, it achieves the best-available convergence rate for non-PL objectives simultaneously while outperforming existing algorithms for PL objectives.
自适应是现代优化理论中一个重要的但尚未得到充分研究的性质。最先进的理论与当前实践之间的差距是惊人的，因为具有理想理论保证的算法通常涉及不同制度下超参数的完全不同设置，例如步长方案和批大小。尽管有吸引人的理论结果，这种分裂的策略提供了很少的洞察力，如果有的话，从业者选择广泛工作的算法而不调整超参数。在这项工作中，我们将Lei ＆ Jordan 2016引入的“几何化”技术和Nguyen et al.， 2017的texttt{莎拉}算法相结合，提出了用于非凸有限和随机优化的几何化texttt{莎拉}算法。我们的算法被证明对目标精度的大小和Polyak-Łojasiewicz (PL)常数都有自适应性。此外，它在非PL目标的同时实现了最佳可用收敛率，同时优于现有的PL目标算法。
{"title":"Adaptivity of Stochastic Gradient Methods for Nonconvex Optimization","authors":"Samuel Horváth, Lihua Lei, Peter Richtárik, Michael I. Jordan","doi":"10.1137/21m1394308","DOIUrl":"https://doi.org/10.1137/21m1394308","url":null,"abstract":"Adaptivity is an important yet under-studied property in modern optimization theory. The gap between the state-of-the-art theory and the current practice is striking in that algorithms with desirable theoretical guarantees typically involve drastically different settings of hyperparameters, such as step-size schemes and batch sizes, in different regimes. Despite the appealing theoretical results, such divisive strategies provide little, if any, insight to practitioners to select algorithms that work broadly without tweaking the hyperparameters. In this work, blending the \"geometrization\" technique introduced by Lei & Jordan 2016 and the texttt{SARAH} algorithm of Nguyen et al., 2017, we propose the Geometrized texttt{SARAH} algorithm for non-convex finite-sum and stochastic optimization. Our algorithm is proved to achieve adaptivity to both the magnitude of the target accuracy and the Polyak-Łojasiewicz (PL) constant if present. In addition, it achieves the best-available convergence rate for non-PL objectives simultaneously while outperforming existing algorithms for PL objectives.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"236 1","pages":"634-648"},"PeriodicalIF":0.0,"publicationDate":"2020-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76432428","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Multi-agent reinforcement learning (MARL), despite its popularity and empirical success, suffers from the curse of dimensionality. This paper builds the mathematical framework to approximate cooperative MARL by a mean-field control (MFC) framework, and shows that the approximation error is of $O(frac{1}{sqrt{N}})$. By establishing appropriate form of the dynamic programming principle for both the value function and the Q function, it proposes a model-free kernel-based Q-learning algorithm (MFC-K-Q), which is shown to be of linear convergence rate, the first of its kind in the MARL literature. It further establishes that the convergence rate and the sample complexity of MFC-K-Q are independent of the number of agents $N$. Empirical studies for the network traffic congestion problem demonstrate that MFC-K-Q outperforms existing MARL algorithms when $N$ is large, for instance when $N>50$.
{"title":"Mean-Field Controls with Q-Learning for Cooperative MARL: Convergence and Complexity Analysis","authors":"Haotian Gu, Xin Guo, Xiaoli Wei, Renyuan Xu","doi":"10.1137/20m1360700","DOIUrl":"https://doi.org/10.1137/20m1360700","url":null,"abstract":"Multi-agent reinforcement learning (MARL), despite its popularity and empirical success, suffers from the curse of dimensionality. This paper builds the mathematical framework to approximate cooperative MARL by a mean-field control (MFC) framework, and shows that the approximation error is of $O(frac{1}{sqrt{N}})$. By establishing appropriate form of the dynamic programming principle for both the value function and the Q function, it proposes a model-free kernel-based Q-learning algorithm (MFC-K-Q), which is shown to be of linear convergence rate, the first of its kind in the MARL literature. It further establishes that the convergence rate and the sample complexity of MFC-K-Q are independent of the number of agents $N$. Empirical studies for the network traffic congestion problem demonstrate that MFC-K-Q outperforms existing MARL algorithms when $N$ is large, for instance when $N>50$.","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"20 1","pages":"1168-1196"},"PeriodicalIF":0.0,"publicationDate":"2020-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74556016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01Epub Date: 2020-05-19DOI: 10.1137/19m1272226
Zixuan Cang, Guo-Wei Wei
Persistent homology is a powerful tool for characterizing the topology of a data set at various geometric scales. When applied to the description of molecular structures, persistent homology can capture the multiscale geometric features and reveal certain interaction patterns in terms of topological invariants. However, in addition to the geometric information, there is a wide variety of nongeometric information of molecular structures, such as element types, atomic partial charges, atomic pairwise interactions, and electrostatic potential functions, that is not described by persistent homology. Although element-specific homology and electrostatic persistent homology can encode some nongeometric information into geometry based topological invariants, it is desirable to have a mathematical paradigm to systematically embed both geometric and nongeometric information, i.e., multicomponent heterogeneous information, into unified topological representations. To this end, we propose a persistent cohomology based framework for the enriched representation of data. In our framework, nongeometric information can either be distributed globally or reside locally on the datasets in the geometric sense and can be properly defined on topological spaces, i.e., simplicial complexes. Using the proposed persistent cohomology based framework, enriched barcodes are extracted from datasets to represent heterogeneous information. We consider a variety of datasets to validate the present formulation and illustrate the usefulness of the proposed method based on persistent cohomology. It is found that the proposed framework outperforms or at least matches the state-of-the-art methods in the protein-ligand binding affinity prediction from massive biomolecular datasets without resorting to any deep learning formulation.
{"title":"Persistent Cohomology for Data With Multicomponent Heterogeneous Information.","authors":"Zixuan Cang, Guo-Wei Wei","doi":"10.1137/19m1272226","DOIUrl":"https://doi.org/10.1137/19m1272226","url":null,"abstract":"<p><p>Persistent homology is a powerful tool for characterizing the topology of a data set at various geometric scales. When applied to the description of molecular structures, persistent homology can capture the multiscale geometric features and reveal certain interaction patterns in terms of topological invariants. However, in addition to the geometric information, there is a wide variety of nongeometric information of molecular structures, such as element types, atomic partial charges, atomic pairwise interactions, and electrostatic potential functions, that is not described by persistent homology. Although element-specific homology and electrostatic persistent homology can encode some nongeometric information into geometry based topological invariants, it is desirable to have a mathematical paradigm to systematically embed both geometric and nongeometric information, i.e., multicomponent heterogeneous information, into unified topological representations. To this end, we propose a persistent cohomology based framework for the enriched representation of data. In our framework, nongeometric information can either be distributed globally or reside locally on the datasets in the geometric sense and can be properly defined on topological spaces, i.e., simplicial complexes. Using the proposed persistent cohomology based framework, enriched barcodes are extracted from datasets to represent heterogeneous information. We consider a variety of datasets to validate the present formulation and illustrate the usefulness of the proposed method based on persistent cohomology. It is found that the proposed framework outperforms or at least matches the state-of-the-art methods in the protein-ligand binding affinity prediction from massive biomolecular datasets without resorting to any deep learning formulation.</p>","PeriodicalId":74797,"journal":{"name":"SIAM journal on mathematics of data science","volume":"2 2","pages":"396-418"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1137/19m1272226","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"39150984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：