Stochastic algorithms are well-known for their performance in the era of big data. In this article, we study nonsmooth stochastic Difference-of-Convex functions (DC) programs-the major class of nonconvex stochastic optimization, which have a variety of applications in divers domains, in particular, machine learning. We propose new online stochastic algorithms based on the state-of-the-art DC Algorithm (DCA)-a powerful approach in nonconvex programming framework, in the online context of streaming data continuously generated by some (unknown) source distribution. The new schemes use the stochastic approximations (SAs) principle: deterministic quantities of the standard DCA are replaced by their noisy estimators constructed using newly arriving samples. The convergence analysis of the proposed algorithms is studied intensively with the help of tools from modern convex analysis and martingale theory. Finally, we study several aspects of the proposed algorithms on an important problem in machine learning: the expected problem in principal component analysis (PCA).
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/TNNLS.2022.3213558 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Auto-tuning of filtered proportional-integral-derivative controller for industrial processes under routine operating conditions.
ISA Trans
December 2024
Department of Automation, Xiamen University, Xiamen City 361000, China. Electronic address:
In general, auto-tuning implementation of PID controllers relies on dual controllers, exciting/identification experiments or some prior knowledge on the process and hence the considerable cost on auto-tuning implementation occurs. To deal with such a problem, a new auto-tuning scheme of the FPID controller is developed for minimum variance tasks under routine operating conditions in which the closed-loop system is running without any external excitation other than natural disturbances. This paper reveals that the stochastic disturbance model can be uniquely determined from the first several terms of the impulse response coefficients of the closed-loop system when the precondition on the time delay and the order of the disturbance model is satisfied.
View Article and Find Full Text PDFCParty: Hierarchically Constrained Partition Function of RNA Pseudoknots.
Bioinformatics
December 2024
Department of Biomedical Engineering, University of Alberta.
Motivation: Biologically relevant RNA secondary structures are routinely predicted by efficient dynamic programming algorithms that minimize their free energy. Starting from such algorithms, one can devise partition function algorithms, which enable stochastic perspectives on RNA structure ensembles. As the most prominent example, McCaskill's partition function algorithm is derived from pseudoknot-free energy minimization.
View Article and Find Full Text PDFA new approach to study stochastic epigenetic mutations in sperm methylome of Vietnam war veterans directly exposed to Agent Orange.
Environ Epigenet
November 2024
Medical Genetics Laboratory, Centro Diagnostico Italiano, Milan, MI 20147, Italy.
Among the various environmental pollutants, dioxin, a highly toxic and widely used compound, is associated with numerous adverse health effects, including a potentially toxic multigenerational effect. Understanding the mechanisms by which dioxin exposure can affect sperm epigenetics is critical to comprehending the potential consequences for offspring health and development. This study investigates the possible association between weighted epimutations, hypothesized as markers of epigenetic drift, and dioxin exposure in sperm tissues.
View Article and Find Full Text PDFLooking into the future: a machine learning powered prediction model for oocyte return rates after cryopreservation.
Reprod Biomed Online
August 2024
Boston IVF - IVIRMA Global Research Alliance, Waltham, MA, USA.
Research Question: Could a predictive model, using data from all US fertility clinics reporting to the Society for Assisted Reproductive Technology, estimate the likelihood of patients using their stored oocytes?
Design: Multiple learner algorithms, including penalized regressions, random forests, gradient boosting machine, linear discriminant analysis and bootstrap aggregating decision trees were used. Data were split into training and test datasets. Patient demographics, medical and fertility diagnoses, partner information and geographic locations were analysed.
Environmental dispersal and host priority effect alternatively dominate intestinal microbiota succession of cultured shrimp along with host development.
Mar Life Sci Technol
November 2024
State Key Laboratory of Biocontrol, School of Marine Sciences, Sun Yat-sen University, Guangzhou, 510275 China.
Unlabelled: Given the importance of the intestinal microbiota in life-long health, increasing attention has been paid to ecological mechanisms that govern microbial succession. Both environmental dispersal and host priority effect play important roles in intestinal microbiota succession of aquatic animals, but their relative importance is unknown. Here, we explore the intestinal microbiota succession and assembly of across larvae, postlarvae, juvenile, and preadult stages.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!