Multi-Task Learning Improves Disease Models from Web Search

Abstract

We investigate the utility of multi-task learning to disease surveillance using Web search data. Our motivation is two-fold. Firstly, we assess whether concurrently training models for various geographies -inside a country or across different countries -can improve accuracy. We also test the ability of such models to assist health systems that are producing sporadic disease surveillance reports that reduce the quantity of available training data. We explore both linear and nonlinear models, specifically a multi-task expansion of elastic net and a multi-task Gaussian Process, and compare them to their respective single task formulations. We use influenza-like illness as a case study and conduct experiments on the United States (US) as well as England, where both health and Google search data were obtained. Our empirical results indicate that multi-task learning improves regional as well as national models for the US. The percentage of improvement on mean absolute error increases up to 14.8% as the historical training data is reduced from 5 to 1 year(s), illustrating that accurate models can be obtained, even by training on relatively short time intervals. Furthermore, in simulated scenarios, where only a few health reports (training data) are available, we show that multi-task learning helps to maintain a stable performance across all the affected locations. Finally, we present results from a cross-country experiment, where data from the US improves the estimates for England. As the historical training data for England is reduced, the benefits of multi-task learning increase, reducing mean absolute error by up to 40%. CCS CONCEPTS•

Transfer Learning for Unsupervised Influenza-like Illness Models from Online Search Data

“…queries with frequencies that are correlated to ILI rates, but have no link to flu, such as 'college basketball' or 'spring break'. To remove these unrelated queries we applied a semantic filter based on word embedding representations, similar to the one proposed in [38,72,73]. Word embeddings were trained on the English Wikipedia corpus using the fastText method [12].…”

“…4 is given by [4,23]. A query's embedding is defined as the average of the embeddings of its tokens, an effective practice for short texts [8,42,66,72]. We denote with v S i , v T j both ∈ R 1×d , the embeddings of a source query (from Q S ) and of a target query from P T , respectively.…”

Privacy-Preserving Crowd-Sourcing of Web Searches with Private Data Donor

“…The former is a Web interface offering a global, longitudinal view of the interest for one or more queries, provided as a normalized number between 0 and 100. The latter is a service offered to researchers, with access to the interest of search queries related to health topics (see also [23,46]). Here, the interest is provided as a frequency, i.e., the number of times the queries were searched during a time period normalized by the total number of searches in the same period.…”

Improved state-level influenza activity nowcasting in the United States leveraging Internet-based data sources and network approaches via ARGONet

“…Our work focuses on a methodology aimed at achieving this for influenza activity 12 surveillance. 13 Influenza has a large seasonal burden across the United States, infecting up to 35 million people 14 and causing between 12000 and 56000 deaths per year [3]. Limiting the spread of outbreaks and reducing 15 morbidity in those already infected are crucial steps for mitigating the impact of influenza.…”

“…Many utilize innovative web-based data sources such as Internet search 31 frequencies and electronic health records [5]. Some have also taken into account historically-observed spatial 32 and temporal synchronicities in flu activity [11,12] to improve the accuracy of existing flu surveillance 33 tools [13,14]. Because influenza transmission occurs locally and is spread from person to person, the timing 34 of outbreaks and resulting infection rate curves can significantly differ from state to state.…”