[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-290922-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-290922-en":131},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","symphony-in-the-latent-space-provably-integrating-high-dimensional-techniques-with-non-linear-machine-learning-models","Symphony in the Latent Space - Provably Integrating High-dimensional Techniques with Non-linear Machine Learning Models","","Revisits machine learning for interacting multivariate time series, such as cross-entity interactions in actively managed portfolios or user interactions in social networks. Aims to forecast future evolution of ensembles by addressing high-dimensional interactions and non-linearity, where ad-hoc integrations often show weak efficacy in time-evolving systems. Proposes an additive influence model that decouples learning of high-dimensional interactions via kernel embeddings in a latent space from learning non-linear feature-response interactions, achieving stronger forecasting on two applications.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/symphony-in-the-latent-space-provably-integrating-high-dimensional-techniques-with-non-linear-machine-learning-models/290922/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/symphony-in-the-latent-space-provably-integrating-high-dimensional-techniques-with-non-linear-machine-learning-models/290922.png","ImageObject",300,407,{"name":42,"@type":43},"Eden","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-31","2026-09-17",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",12,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What problem does the paper target?","Question",{"text":63,"@type":64},"It targets building ML algorithms for systems where multiple entities interact over time, aiming to forecast future behavior of an ensemble of multivariate time series using information available up to time t.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"Why is designing such forecasting models challenging?",{"text":68,"@type":64},"Two main challenges are emphasized: cross-entity interaction leads to high-dimensionality, and feature-response relationships are non-linear, making linear models insufficient.",{"name":70,"@type":61,"acceptedAnswer":71},"What is the proposed framework and what does it achieve?",{"text":72,"@type":64},"The paper proposes the additive influence model, which decouples learning high-dimensional interactions from learning non-linear feature-response interactions. It uses kernel-based techniques for provable latent-space embeddings and introduces statistically sound non-parametric and ensemble methods, yielding significantly stronger forecasting power in experiments.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},290922,1789995989,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,111,115,120,123,127],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":25,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":112,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":113,"slug":114},8,30,"research-report",{"id":116,"doc_module":4,"doc_module_name":25,"category_name":117,"show_sort_weight":118,"slug":119},9,"Religion & Spirituality",20,"religion-spirituality",{"id":118,"doc_module":4,"doc_module_name":25,"category_name":121,"show_sort_weight":118,"slug":122},"World Cup","world-cup",{"id":124,"doc_module":4,"doc_module_name":25,"category_name":125,"show_sort_weight":124,"slug":126},10,"Lifestyle","lifestyle",{"id":128,"doc_module":4,"doc_module_name":25,"category_name":129,"show_sort_weight":97,"slug":130},19,"General","general",{"code":4,"msg":82,"data":132},{"doc_id":79,"user_id":133,"nickname":42,"user_avatar":134,"doc_module":4,"category_id":112,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":140,"language":141,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":12,"update_tm":145,"read_time":146,"preview_page_count":4},1374391974468,"https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0","Symphony in the Latent Space: Provably Integrating High-dimensional Techniques  \nwith Non-linear Machine Learning Models  \nQiong Wu* 1 , Jian Li2 , Zhenming Liu1 , Yanhua Li3 , Mihai Cucuringu4  \n1William & Mary  \n2 Tsinghua University  \n2Worcester Polytechnic Institute  \n4University of Oxford and The Alan Turing Institute  \nAbstract  \nThis paper revisits building machine learning algorithms that involve interactions between entities, such as those between financial assets in an actively managed portfolio, or interactions between users in a social network. Our goal is to forecast the future evolution of ensembles of multivariate time series in such applications (e.g., the future return of a financial asset or the future popularity of a Twitter account) . Designing ML algorithms for such systems requires addressing the challenges of high-dimensional interactions and non-linearity. Existing approaches usually adopt an ad-hoc approach to integrating high-dimensional techniques into non-linear models and recent studies have shown these approaches have questionable efficacy in time-evolving interacting systems.  \nTo this end, we propose a novel framework, which we dub asthe additive influence model. Under our modeling assumption, we show that it is possible to decouple the learning of high-dimensional interactions from the learning of non-linear feature interactions. To learn the high-dimensional interactions, we leverage kernel-based techniques, with provable guarantees, to embed the entities in a low-dimensional latent space. To learn the non-linear feature-response interactions, we generalize prominent machine learning techniques, including designing a new statistically sound non-parametric method and an ensemble learning algorithm optimized for vector regressions. Extensive experiments on two common applications demonstrate that our new algorithms deliver significantly stronger forecasting power compared to standard and recently proposed methods.  \nIntroduction  \nWe revisit the problem of building machine learning algorithms that involve interactions between entities, such as those between users and items in a recommendation system, or between financial assets in an actively managed portfolio, or between populations in different counties in a diseasespreading process. Our proposed forecasting model uses information available up to time t to predict yt+1,i, the future behavior of entity i at time t + 1 (e.g., the future price of stock i at time t + 1), for a total number of d entities (Laptev et al. 2017; Farhangi et al. 2022) . Designing such models has  \n* Currently working at AT&T Labs.  \nCopyright © 2023, Association for the Advancement of Artificial Intelligence ([www.aaai.org](www.aaai.org)). All rights reserved.  \nproven remarkably difficult, as one needs to circumvent two main challenges that require often incompatible solutions.  \n1. Cross-entity interaction: high-dimensionality. In many ensembles of multivariate time series systems, it is often the case that the current state of one entity could potentially impact the future state of another. When considering the equity market as an example, Amazon’s disclosure of its revenue change in cloud services could indicate that the revenues of other cloud providers (e.g., competitors) could also change.  \nThe interaction is high-dimensional because the total possible number of interactions is usually much larger than the number of available observations. For example, in a portfolio of 3,000 stocks, the total number of potential links between pairs of stocks is 3 , 000 × 3 , 000 ≈ 107 , but we often have only 2,500 data points (e.g., 10 years of daily data), and thus capturing the cross-entity interactions becomes a very challenging problem.  \n2. Feature-response interactions: non-linearity. Linear models are usually insufficient to characterize the relationship between the response/label and the available information (features), thus techniques beyond simple linear regressions","cbCaimcT0YHec5rv","https://ap.wps.com/l/cbCaimcT0YHec5rv","pdf",1860339,26,"English","# Abstract\n# Introduction\n## Cross-entity interaction: high-dimensionality\n## Feature-response interactions: non-linearity\n## Provable cross-entity models (CEM)\n## Practical univariate models (UM)","[{\"question\":\"What problem does the paper target?\",\"answer\":\"It targets building ML algorithms for systems where multiple entities interact over time, aiming to forecast future behavior of an ensemble of multivariate time series using information available up to time t.\"},{\"question\":\"Why is designing such forecasting models challenging?\",\"answer\":\"Two main challenges are emphasized: cross-entity interaction leads to high-dimensionality, and feature-response relationships are non-linear, making linear models insufficient.\"},{\"question\":\"What is the proposed framework and what does it achieve?\",\"answer\":\"The paper proposes the additive influence model, which decouples learning high-dimensional interactions from learning non-linear feature-response interactions. It uses kernel-based techniques for provable latent-space embeddings and introduces statistically sound non-parametric and ensemble methods, yielding significantly stronger forecasting power in experiments.\"}]","Symphony in the Latent Space - Provably Integrating High-dimensional Techniques with Non-linear Machine Learning Models | PDF",1789642958,66]