English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

In high stakes applications such as healthcare and finance analytics, theinterpretability of predictive models is required and necessary for domainpractitioners to trust the predictions. Traditional machine learning models,e.g., logistic regression (LR), are easy to interpret in nature. However, manyof these models aggregate time-series data without considering the temporalcorrelations and variations. Therefore, their performance cannot match up torecurrent neural network (RNN) based models, which are nonetheless difficult tointerpret. In this paper, we propose a general framework TRACER to facilitateaccurate and interpretable predictions, with a novel model TITV devised forhealthcare analytics and other high stakes applications such as financialinvestment and risk management. Different from LR and other existing RNN-basedmodels, TITV is designed to capture both the time-invariant and thetime-variant feature importance using a feature-wise transformation subnetworkand a self-attention subnetwork, for the feature influence shared over theentire time series and the time-related importance respectively. Healthcareanalytics is adopted as a driving use case, and we note that the proposedTRACER is also applicable to other domains, e.g., fintech. We evaluate theaccuracy of TRACER extensively in two real-world hospital datasets, and ourdoctors/clinicians further validate the interpretability of TRACER in both thepatient level and the feature level. Besides, TRACER is also validated in ahigh stakes financial application and a critical temperature forecastingapplication. The experimental results confirm that TRACER facilitates bothaccurate and interpretable analytics for high stakes applications.

TRACER: A Framework for Facilitating Accurate and Interpretable Analytics for High Stakes Applications