Learning Hierarchical Representations of Electronic Health Records for Clinical Outcome Prediction.

Clinical outcome prediction based on Electronic Health Record (EHR) helps enable early interventions for high-risk patients, and is thus a central task for smart healthcare. Conventional deep sequential models fail to capture the rich temporal patterns encoded in the long and irregular clinical event sequences in EHR. We make the observation that clinical events at a long time scale exhibit strong temporal patterns, while events within a short time period tend to be disordered co-occurrence. We thus propose differentiated mechanisms to model clinical events at different time scales. Our model learns hierarchical representations of event sequences, to adaptively distinguish between short-range and long-range events, and accurately capture their core temporal dependencies. Experimental results on real clinical data show that our model greatly improves over previous state-of-the-art models, achieving AUC scores of 0.94 and 0.90 for predicting death and ICU admission, respectively. Our model also successfully identifies important events for different clinical outcome prediction tasks.

Thickness-controlled direct growth of nanographene and nanographite film on non-catalytic substrates.

Metal-catalyzed chemical vapor deposition (CVD) has been broadly employed for large-scale production of high-quality graphene. However, a following transfer process to targeted substrates is needed, which is incompatible with current silicon technology. We here report a new CVD approach to form nanographene and nanographite films with accurate thickness control directly on non-catalytic substrates such as silicon dioxide and quartz at 800 °C. The growth time is as short as a few seconds. The approach includes using 9-bis(diethylamino)silylanthracene as the carbon source and an atomic layer deposition (ALD) controlling system. The structure of the formed nanographene and nanographite films were characterized using atomic force microscopy, high resolution transmission electron microscopy, Raman scattering, and x-ray photoemission spectroscopy. The nanographite film exhibits a transmittance higher than 80% at 550 nm and a sheet electrical resistance of 2000 ohms per square at room temperature. A negative temperature-dependence of the resistance of the nanographite film is also observed. Moreover, the thickness of the films can be precisely controlled via the deposition cycles using an ALD system, which promotes great application potential for optoelectronic and thermoelectronic-devices.

Learning Scalable Deep Kernels with Recurrent Structure.

Many applications in speech, robotics, finance, and biology deal with sequential data, where ordering matters and recurrent structures are common. However, this structure cannot be easily captured by standard kernel functions. To model such structure, we propose expressive closed-form kernel functions for Gaussian processes. The resulting model, GP-LSTM, fully encapsulates the inductive biases of long short-term memory (LSTM) recurrent networks, while retaining the non-parametric probabilistic advantages of Gaussian processes. We learn the properties of the proposed kernels by optimizing the Gaussian process marginal likelihood using a new provably convergent semi-stochastic gradient procedure, and exploit the structure of these kernels for scalable training and prediction. This approach provides a practical representation for Bayesian LSTMs. We demonstrate state-of-the-art performance on several benchmarks, and thoroughly investigate a consequential autonomous driving application, where the predictive uncertainties provided by GP-LSTM are uniquely valuable.