Unifying Fourteen Post-Hoc Attribution Methods With Taylor Interactions.

Various attribution methods have been developed to explain deep neural networks (DNNs) by inferring the attribution/importance/contribution score of each input variable to the final output. However, existing attribution methods are often built upon different heuristics. There remains a lack of a unified theoretical understanding of why these methods are effective and how they are related.

A Hybrid Residual Network and Long Short-Term Memory Method for Peptic Ulcer Bleeding Mortality Prediction.

The prediction of patient mortality, which can detect high-risk patients, is a significant yet challenging problem in medical informatics. Thanks to the wide adoption of electronic health records (EHRs), many data-driven methods have been proposed to forecast mortality. However, most existing methods do not consider correlations between static and dynamic data, which contain significant information about mutual influences between these data.

Machine learning model to predict recurrent ulcer bleeding in patients with history of idiopathic gastroduodenal ulcer bleeding.

Background: Patients with a history of Helicobacter pylori-negative idiopathic bleeding ulcers have an increased risk of recurring ulcer complications.

Aim: To build a machine learning model to identify patients at high risk for recurrent ulcer bleeding.

Methods: Data from a retrospective cohort of 22 854 patients (training cohort) diagnosed with peptic ulcer disease in 2007-2016 were analysed to build a model (IPU-ML) to predict recurrent ulcer bleeding.

Mesoporous zinc ferrite: synthesis, characterization, and photocatalytic activity with H2O2/visible light.

Mesoporous ZnFe(2)O(4) (meso-ZnFe(2)O(4)) was synthesized by a hydrothermal process in which cetyltrimethylammonium bromide (CTAB) participates in the reaction to produce nanocrystals. Synthesized ZnFe(2)O(4) was characterized by energy dispersive spectroscopy (EDS), X-ray diffraction (XRD), Brunauer-Emmett-Teller (BET) surface area, scanning electronic microscopy (SEM), transmission electron microscopy (TEM), and diffuse reflectance spectra (DRS). The meso-ZnFe(2)O(4) was resulted from the agglomeration of nanoparticles with size of 5-10nm.