The Clinical Language Understanding group at Nuance Communications has developed a medical information extraction system that combines a rule-based extraction engine with machine learning algorithms to identify and categorize references to patient smoking in clinical reports. The extraction engine identifies smoking references; documents that contain no smoking references are classified as UNKNOWN. For the remaining documents, the extraction engine uses linguistic analysis to associate features such as status and time to smoking mentions. Machine learning is used to classify the documents based on these features. This approach shows overall accuracy in the 90s on all data sets used. Classification using engine-generated and word-based features outperforms classification using only word-based features for all data sets, although the difference gets smaller as the data set size increases. These techniques could be applied to identify other risk factors, such as drug and alcohol use, or a family history of a disease.
Generalized additive models (GAMs) are favored in many regression and binary classification problems because they are able to fit complex, nonlinear functions while still remaining interpretable. In the first part of this paper, we generalize a state-of-the-art GAM learning algorithm based on boosted trees to the multiclass setting, showing that this multiclass algorithm outperforms existing GAM learning algorithms and sometimes matches the performance of full complexity models such as gradient boosted trees.In the second part, we turn our attention to the interpretability of GAMs in the multiclass setting. Surprisingly, the natural interpretability of GAMs breaks down when there are more than two classes. Naive interpretation of multiclass GAMs can lead to false conclusions. Inspired by binary GAMs, we identify two axioms that any additive model must satisfy in order to not be visually misleading. We then develop a technique called Additive Post-Processing for Interpretability (API) that provably transforms a pretrained additive model to satisfy the interpretability axioms without sacrificing accuracy. The technique works not just on models trained with our learning algorithm, but on any multiclass additive model, including multiclass linear and logistic regression. We demonstrate the effectiveness of API on a 12-class infant mortality dataset.Interpretable models, though sometimes less accurate than blackbox models, are preferred in many real-world applications. In criminal justice, finance, hiring, and other domains that impact people's lives, interpretable models are often used because their transparency helps determine if a model is biased or unsafe [26,31].
Graphs are ubiquitous and are often used to understand the dynamics of a system. Probabilistic Graphical Models comprising Bayesian and Markov networks, and Conditional Independence graphs are some of the popular graph representation techniques. They can model relationships between features (nodes) together with the underlying distribution. Although theoretically these models can represent very complex dependency functions, in practice often simplifying assumptions are made due to computational limitations associated with graph operations. This work introduces Neural Graphical Models (NGMs) which attempt to represent complex feature dependencies with reasonable computational costs. Specifically, given a graph of feature relationships and corresponding samples, we capture the dependency structure between the features along with their complex function representations by using neural networks as a multi-task learning framework. We provide efficient learning, inference and sampling algorithms for NGMs. Moreover, NGMs can fit generic graph structures including directed, undirected and mixededge graphs as well as support mixed input data types. We present empirical studies that show NGMs' capability to represent Gaussian graphical models, inference analysis of a lung cancer data and extract insights from a real world infant mortality data provided by CDC.
Sudden unexpected infant death (SUID) is a broad term that describes the death of an infant (<365 days of age) that occurs suddenly and unexpectedly, and the cause is not obvious before investigation. It includes three causes of death as classified by the International Classification of Diseases, 10th Revision (ICD-10): sudden infant death syndrome (SIDS, R95), ill-defined causes (R99) and accidental suffocation and strangulation in bed (W75) which, combined, result in approximately 3700 deaths annually in the United States. 1 The age distribution of SUID is well described and unique, with relatively small numbers in the first month of life, a peak at 2-3 months and approximately 90% of deaths occurring before 6 months of age. This age distribution is often referred to as a critical developmental age. 2 While this is the least studied of the three SIDS risk areas proposed in Filiano and Kinney's Triple Risk
Probabilistic Graphical Models (PGMs) are generative models of complex systems. They rely on conditional independence assumptions between variables to learn sparse representations which can be visualized in a form of a graph. Such models are used for domain exploration and structure discovery in poorly understood domains. This work introduces a novel technique to perform sparse graph recovery by optimizing deep unrolled networks. Assuming that the input data X ∈ R M ×D comes from an underlying multivariate Gaussian distribution, we apply a deep model on X that outputs the precision matrix Θ, which can also be interpreted as the adjacency matrix. Our model, uGLAD 1 , builds upon and extends the state-ofthe-art model GLAD [42] to the unsupervised setting. The key benefits of our model are (1) uGLAD automatically optimizes sparsity-related regularization parameters leading to better performance than existing algorithms. (2) We introduce multitask learning based 'consensus' strategy for robust handling of missing data in an unsupervised setting. We evaluate model results on synthetic Gaussian data, non-Gaussian data generated from Gene Regulatory Networks, and present a case study in anaerobic digestion.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.