Simultaneous Parameter Learning and Bi-Clustering for Multi-Response Models

Published on Aug 14, 2019
路 DOI :10.3389/fdata.2019.00027
Ming Yu3
Estimated H-index: 3
(U of C: University of Chicago),
Karthikeyan Natesan Ramamurthy14
Estimated H-index: 14
+ 1 AuthorsAurelie C. Lozano15
Estimated H-index: 15
We consider multi-response and multi-task regression models, where the parameter matrix to be estimated is expected to have an unknown grouping structure. The groupings can be along tasks, or features, or both, the last one indicating a bi-cluster or "checkerboard" structure. Discovering this grouping structure along with parameter inference makes sense in several applications, such as multi-response Genome-Wide Association Studies (GWAS). By inferring this additional structure we can obtain valuable information on the underlying data mechanisms (e.g. relationships among genotypes and phenotypes in GWAS). In this paper, we propose two formulations to simultaneously learn the parameter matrix and its group structures, based on convex regularization penalties. We present optimization approaches to solve the resulting problems and provide numerical convergence guarantees. Extensive experiments demonstrate much better clustering quality compared to other methods, and our approaches are also validated on real datasets concerning phenotypes and genotypes of plant varieties.
  • References (10)
  • Citations (0)
馃摉 Papers frequently viewed together
2012WABI: Workshop on Algorithms in Bioinformatics
5 Authors (Yi Shi, ..., Dale Schuurmans)
2 Citations
7 Citations
78% of Scinapse members use related papers. After signing in, all features are FREE.
#1Josie Hilley (A&M: Texas A&M University)H-Index: 5
#2Brock D. Weers (A&M: Texas A&M University)H-Index: 6
Last. John E. Mullet (A&M: Texas A&M University)H-Index: 76
view all 8 authors...
Sorghum is an important C4 grass crop grown for grain, forage, sugar, and bioenergy production. While tall, late flowering landraces are commonly grown in Africa, short early flowering varieties were selected in US grain sorghum breeding programs to reduce lodging and to facilitate machine harvesting. Four loci have been identified that affect stem length (Dw1-Dw4). Subsequent research showed that Dw3 encodes an ABCB1 auxin transporter and Dw1 encodes a highly conserved protein involved in the r...
13 CitationsSource
#1Eric C. Chi (NCSU: North Carolina State University)H-Index: 10
#2Genevera I. Allen (Rice University)H-Index: 20
Last. Richard G. Baraniuk (Rice University)H-Index: 87
view all 3 authors...
41 CitationsSource
#1Josie Hilley (A&M: Texas A&M University)H-Index: 5
#2Sandra K. Truong (A&M: Texas A&M University)H-Index: 9
Last. John E. Mullet (A&M: Texas A&M University)H-Index: 76
view all 5 authors...
Sorghum is an important C4 grain and grass crop used for food, feed, forage, sugar, and biofuels. In its native Africa, sorghum landraces often grow to approximately 3鈥4 meters in height. Following introduction into the U.S., shorter, early flowering varieties were identified and used for production of grain. Quinby and Karper identified allelic variation at four loci designated Dw1-Dw4 that regulated plant height by altering the length of stem internodes. The current study used a map-based clon...
24 CitationsSource
#1Hanen Borchani (AAU: Aalborg University)H-Index: 7
#2Gherardo Varando (UPM: Technical University of Madrid)H-Index: 4
Last. Pedro L. L贸pez-Cruz (UPM: Technical University of Madrid)H-Index: 44
view all 4 authors...
In recent years, a plethora of approaches have been proposed to deal with the increasingly challenging task of multi-output regression. This study provides a survey on state-of-the-art multi-output regression methods, that are categorized as problem transformation and algorithm adaptation methods. In addition, we present the mostly used performance evaluation measures, publicly available data sets for multi-output regression real-world problems, as well as open-source software frameworks. WIREs ...
106 CitationsSource
Aug 10, 2015 in KDD (Knowledge Discovery and Data Mining)
#1David Hallac (Stanford University)H-Index: 8
#2Jure Leskovec (Stanford University)H-Index: 86
Last. Stephen Boyd (Stanford University)H-Index: 102
view all 3 authors...
Convex optimization is an essential tool for modern data analysis, as it provides a framework to formulate and solve many problems in machine learning and data mining. However, general convex optimization solvers do not scale well, and scalable solvers are often specialized to only work on a narrow class of problems. Therefore, there is a need for simple, scalable algorithms that can solve many common optimization problems. In this paper, we introduce the network lasso, a generalization of the g...
105 CitationsSource
#1Elizabeth D. Schifano (UConn: University of Connecticut)H-Index: 8
#2Lin Li (Harvard University)H-Index: 2
Last. Xihong Lin (Harvard University)H-Index: 55
view all 4 authors...
There is increasing interest in the joint analysis of multiple phenotypes in genome-wide association studies (GWASs), especially for the analysis of multiple secondary phenotypes in case-control studies and in detecting pleiotropic effects. Multiple phenotypes often measure the same underlying trait. By taking advantage of similarity across phenotypes, one could potentially gain statistical power in association analysis. Because continuous phenotypes are likely to be measured on different scales...
54 CitationsSource
#1Nicolai Meinshausen (University of California, Berkeley)H-Index: 25
#2Bin Yu (University of California, Berkeley)H-Index: 51
The Lasso [28] is an attractive technique for regularization and variable selection for high-dimensional data, where the number of predictor variables p is potentially much larger than the number of samples n. However, it was recently discovered [23, 38, 39] that the sparsity pattern of the Lasso estimator can only be asymptotically identical to the true sparsity pattern if the design matrix satisfies the so-called irrepresentable condition. The latter condition can easily be violated in applica...
548 CitationsSource
#1Patrick L. Combettes (University of Paris)H-Index: 47
#2Jean-Christophe Pesquet (Institut Gaspard Monge)H-Index: 39
A broad range of inverse problems can be abstracted into the problem of minimizing the sum of several convex functions in a Hilbert space. We propose a proximal decomposition algorithm for solving this problem with an arbitrary number of nonsmooth functions and establish its weak convergence. The algorithm fully decomposes the problem in that it involves each function individually via its own proximity operator. A significant improvement over the methods currently in use in the area of inverse p...
224 CitationsSource
We consider the generic regularized optimization problem 尾(位) = argmin尾 L(y, X尾) + 位J(尾). Efron, Hastie, Johnstone and Tibshirani [Ann. Statist. 32 (2004) 407-499] have shown that for the LASSO-that is, if L is squared error loss and J(尾) = 鈭ノ测垾 1 is the l 1 norm of 尾-the optimal coefficient path is piecewise linear, that is, 鈭偽(位)/鈭偽. is piecewise constant. We derive a general characterization of the properties of (loss L, penalty J) pairs which give piecewise linear coefficient paths. Such pair...
362 CitationsSource
#1Dilbag S. Multani (DuPont Pioneer)H-Index: 6
#2Steven P. BriggsH-Index: 46
Last. Gurmukh S. Johal (Purdue University)H-Index: 19
view all 6 authors...
Agriculturally advantageous reduction in plant height is usually achieved by blocking the action or production of gibberellins. Here, we describe a different dwarfing mechanism found in maize brachytic2 (br2) mutants characterized by compact lower stalk internodes. The height reduction in these plants results from the loss of a P-glycoprotein that modulates polar auxin transport in the maize stalk. The sorghum ortholog of br2 is dwarf3 (dw3), an unstable mutant of long-standing commercial intere...
290 CitationsSource
Cited By0
#2Bhanukiran Vinzamuri (IBM)H-Index: 6
Last. Amit Dhurandhar (IBM)H-Index: 7
view all 4 authors...
In recent years, post-hoc local instance-level and global dataset-level explainability of black-box models has received a lot of attention. Much less attention has been given to obtaining insights at intermediate or group levels, which is a need outlined in recent works that study the challenges in realizing the guidelines in the General Data Protection Regulation (GDPR). In this paper, we propose a meta-method that, given a typical local explainability method, can build a multilevel explanation...