This paper processes and combines an extensive collection of publicly available datasets to provide a unified information source for representing geographical regions with regards to their pandemic-related behavior. The features are grouped into various categories to account for their impact based on the higher-level concepts associated with them. This work uses several correlation analysis techniques to observe value and order relationships between features, feature groups, and COVID-19 occurrences. Dimensionality reduction techniques and projection methodologies are used to elaborate on individual and group importance of these representative features. In addition, a specific RNN-based inference pipeline called DoubleWindowLSTM-CP is designed in this work for predictive event modeling, thus utilizing sequential patterns as well as enabling concise record representation.
Despite the success of deep learning in domains such as image, voice, and graphs, there has been little progress in deep representation learning for domains without a known structure between features. For instance, a tabular dataset of different demographic and clinical factors where the feature interactions are not given as a prior. In this paper, we propose Group-Connected Multilayer Perceptron (GMLP) networks to enable deep representation learning in these domains. GMLP is based on the idea of learning expressive feature combinations (groups) and exploiting them to reduce the network complexity by defining local group-wise operations. During the training phase, GMLP learns a sparse feature grouping matrix using temperature annealing softmax with an added entropy loss term to encourage the sparsity. Furthermore, an architecture is suggested which resembles binary trees, where group-wise operations are followed by pooling operations to combine information; reducing the number of groups as the network grows in depth. To evaluate the proposed method, we conducted experiments on different real-world datasets covering various application areas. Additionally, we provide visualizations on MNIST and synthesized data. According to the results, GMLP is able to successfully learn and exploit expressive feature combinations and achieve state-of-the-art classification performance on different datasets.tincidunt magna sed ex sollicitudin condimentum.
Effective representation learning of electronic health records is a challenging task and is becoming more important as the availability of such data is becoming pervasive. The data contained in these records are irregular and contain multiple modalities such as notes, and medical codes. They are preempted by medical conditions the patient may have, and are typically recorded by medical staff. Accompanying codes are notes containing valuable information about patients beyond the structured information contained in electronic health records. We use transformer networks and the recently proposed BERT language model to embed these data streams into a unified vector representation. The presented approach effectively encodes a patient's visit data into a single distributed representation, which can be used for downstream tasks. Our model demonstrates superior performance and generalization on mortality, readmission and length of stay tasks using the publicly available MIMIC-III ICU dataset.
In this paper, we study the problem of children activity recognition using smartwatch devices. We introduce the need for a robust children activity model and challenges involved. To address the problem, we employ two deep neural network models, specifically, Bi-Directional LSTM model and a fully connected deep network and compare the results to commonly used models in the area. We demonstrate that our proposed deep models can significantly improve results compared to baseline models. We further show benefits of activity intensity level detection in health monitoring and verify high performance of our proposed models in this task.