Recommender systems have been widely adopted by electronic commerce and entertainment industries for individualized prediction recommendation, which benefit consumers improve business intelligence. In this article, we propose an innovative method, namely the recommendation engine of multilayers (REM), tensor recommender systems. The proposed method utilizes structure a response to integrate information from multiple modes, creates additional layer nested latent factors accommodate...

Because of the accessibility big data collections from consumers, products, and stores, advanced sales forecasting capabilities have drawn great attention many businesses, especially those in retail, because importance decision making. Improvement accuracy, even by a small percentage, may substantial impact on companies’ production financial planning, marketing strategies, inventory controls, supply chain management. Specifically, our research goal is to forecast each product store near...

This article provides an overview of tensors, their properties, and applications in statistics. Tensors, also known as multidimensional arrays, are generalizations matrices to higher orders useful data representation architectures. We first review basic tensor concepts decompositions, then we elaborate traditional recent tensors the fields recommender systems imaging analysis. illustrate for network explore relations among interacting units a complex system. Some canonical computational...

In recent years, there has been a growing demand to develop efficient recommender systems which track users' preferences and recommend potential items of interest users. this article, we propose group-specific method use dependency information from users share similar characteristics under the singular value decomposition framework. The new approach is effective for "cold-start" problem, where, in testing set, majority responses are obtained or items, their preference not available training...

Heritability is well documented for psychiatric disorders and cognitive abilities which are, however, complex, involving both genetic environmental factors. Hence, it remains challenging to discover how variations contribute such complex traits. In this article, they propose use mediation analysis bridge gap, where neuroimaging phenotypes were utilized as intermediate variables. The Philadelphia Neurodevelopmental Cohort was investigated using genome-wide association studies (GWAS) analyses....

On short video sharing platforms, users often choose background music for their videos. In this paper, we study the problem of recommendation videos on platforms. our setting, item (music) is not recommended directly to user, but created by user. When making recommendations videos, consider three important players: users, and music. We define a unique design novel model address problem. propose based deep learning framework effectively distinctive three-way relationships among Our considers...

This work is motivated by multimodality breast cancer imaging data, which quite challenging in that the signals of discrete tumor-associated microvesicles are randomly distributed with heterogeneous patterns. imposes a significant challenge for conventional regression and dimension reduction models assuming homogeneous feature structure. We develop an innovative multilayer tensor learning method to incorporate heterogeneity higher-order decomposition predict disease status effectively...

Recommender systems have been extensively used by the entertainment industry, business marketing and biomedical industry. In addition to its capacity of providing preference-based recommendations as an unsupervised learning methodology, it has also proven useful in sales forecasting, product introduction other production related businesses. Since some consumers companies need a recommendation or prediction for future budget, labor supply chain coordination, dynamic recommender precise...

Limited access to large-scale data is a key obstacle building machine learning (ML) applications in practice, partly due reluctance of information exchange among owners out privacy and security concerns. To address this “information silo” problem, federated (FL) techniques have been proposed enable decentralized model training via an orchestrating central server received increasing attention several industries (including healthcare finance). Despite its superior protection property, adoption...

Brain-imaging data have been increasingly used to understand intellectual disabilities. Despite significant progress in biomedical research, the mechanisms for most of disabilities remain unknown. Finding underlying neurological has proved difficult, especially children due rapid development their brains. We investigate verbal reasoning, which is a reliable measure individuals' general abilities, and develop class high-order imaging regression models identify brain subregions might be...

Correlation structure contains important information about longitudinal data. Existing sufficient dimension reduction approaches assuming independence may lead to substantial loss of efficiency. We apply the quadratic inference function incorporate correlation and transformation method recover central subspace. The proposed estimators are shown be consistent more efficient than ones independence. In addition, esti- mated subspace is also when taken into account. compare with other through...

We propose a novel multivariate model for analyzing hybrid traits and identifying genetic factors comorbid conditions. Comorbidity is common phenomenon in mental health which an individual suffers from multiple disorders simultaneously. For example, the Study of Addiction: Genetics Environment (SAGE), alcohol nicotine addiction were recorded through assessments that we refer to as traits. Statistical inference studying basis has not been well developed. Recent rank-based methods have...

Understanding the functional mechanisms of complex biological system as a whole is drawing more and attention in global health care management. Traditional Chinese Medicine (TCM), essentially different from Western (WM), gaining increasing due to its emphasis on individual wellness natural herbal medicine, which satisfies goal integrative medicine. However, with explosive growth biomedical data Web, researchers are now confronted problem large-scale analysis query. Besides that, also has...

Bias in adolescent self-reported height and weight is well documented. Given the importance widespread use of National Longitudinal Study Adolescent to Adult Health (Add Health) data for obesity research, we developed tested feasibility validity an empirically derived statistical correction self-report bias wave 1 (W1) Add Health, a large panel study United States. Participants grades 7–12 with complete at W1 were included (n = 20,175). We used measured (SR) relevant biopsychosocial factors...

In this study, we present an application paradigm in which unsupervised machine learning approach is applied to the high-dimensional influenza genetic sequences investigate whether vaccine a driving force evolution of virus. We first used visualization visualize evolutionary paths vaccine-controlled and non-vaccine-controlled viruses low-dimensional space. then quantified differences between their trajectories through use within- between-scatter matrices computation provide statistical...

How to acquire the most valuable consumers grow your recommender system? We propose a dynamic consumer acquisition model enable value-driven decisions. build of that takes into account value contributes system, cost their participation (e.g., privacy loss), and other (via network externality). also data-driven procedures estimate this informed, On three different data sets, we perform comprehensive simulation-based evaluations demonstrate performance model. find nuanced relationships between...

Yu Wanga, Xuan Bib & Annie Quc* a Electrical and Computer Engineering, Duke University, Durham, NC; b Carlson School of Management, University Minnesota, Minneapolis, MN; c Department Statistics, Illinois at Urbana-Champaign, Champaign, IL

Safeguarding intellectual property and preventing potential misuse of AI-generated images are paramount importance. This paper introduces a robust agile plug-and-play watermark detection framework, dubbed as RAW. As departure from traditional encoder-decoder methods, which incorporate fixed binary codes watermarks within latent representations, our approach learnable directly into the original image data. Subsequently, we employ classifier that is jointly trained with to detect presence...

Retrieval-Augmented Generation (RAG) has been empirically shown to enhance the performance of large language models (LLMs) in knowledge-intensive domains such as healthcare, finance, and legal contexts. Given a query, RAG retrieves relevant documents from corpus integrates them into LLMs' generation process. In this study, we investigate adversarial robustness RAG, focusing specifically on examining retrieval system. First, across 225 different setup combinations corpus, retriever, targeted...

Most classification techniques in machine learning are able to produce probability predictions addition class predictions. However, these predicted probabilities often not well calibrated that they deviate from the actual outcome rates (i.e., proportion of data instances actually belong a certain class). A lack calibration can jeopardize downstream decision tasks rely on accurate Although several post hoc methods have been proposed, generally do consider potentially asymmetric costs...

Federated learning (FL) is a privacy-preserving technique that enables distributed computing devices to train shared models across data silos collaboratively. Existing FL works mostly focus on designing advanced algorithms improve the model performance. However, economic considerations of clients, such as fairness and incentive, are yet be fully explored. Without considerations, self-motivated clients may lose interest leave federation. To address this problem, we designed novel incentive...

The growing dependence on machine learning in real-world applications emphasizes the importance of understanding and ensuring its safety. Backdoor attacks pose a significant security risk due to their stealthy nature potentially serious consequences. Such involve embedding triggers within model with intention causing malicious behavior when an active trigger is present while maintaining regular functionality without it. This paper evaluates effectiveness any backdoor attack incorporating...