arXiv subjects

Varun Chandola

Publications and source records attributed to Varun Chandola.

25 records · Page 2Linked to original sources

arXiv2018

Hospital Readmission Prediction - Applying Hierarchical Sparsity Norms for Interpretable Models

Hospital readmissions have become one of the key measures of healthcare quality. Preventable readmissions have been identified as one of the primary targets for reducing costs and improving healthcare delivery. However, most data driven studies for understanding readmissions have produced black box classification and predictive models with moderate performance, which precludes them from being used effectively within the decision support systems in the hospitals. In this paper we present an application of structured sparsity-inducing norms for predicting readmission risk for patients based on their disease history and demographics. Most existing studies have focused on hospital utilization, test results, etc., to assign a readmission label to each episode of hospitalization. However, we focus on assigning a readmission risk label to a patient based on their disease history. Our emphasis is on interpreting the models to improve the understanding of the readmission problem. To achieve this, we exploit the domain induced hierarchical structure available for the disease codes which are the features for the classification algorithm. We use a tree based sparsity-inducing regularization strategy that explicitly uses the domain hierarchy. The resulting model not only outperforms standard regularization procedures but is also highly sparse and interpretable. We analyze the model and identify several significant factors that have an effect on readmission risk. Some of these factors conform to existing beliefs, e.g., impact of surgical complications and infections during hospital stay. Other factors, such as the impact of mental disorder and substance abuse on readmission, provide empirical evidence for several pre-existing but unverified hypotheses. The analysis also reveals previously undiscovered connections such as the influence of socioeconomic factors like lack of housing and malnutrition.

cs.LG↗

arXiv2018

S-Isomap++: Multi Manifold Learning from Streaming Data

Manifold learning based methods have been widely used for non-linear dimensionality reduction (NLDR). However, in many practical settings, the need to process streaming data is a challenge for such methods, owing to the high computational complexity involved. Moreover, most methods operate under the assumption that the input data is sampled from a single manifold, embedded in a high dimensional space. We propose a method for streaming NLDR when the observed data is either sampled from multiple manifolds or irregularly sampled from a single manifold. We show that existing NLDR methods, such as Isomap, fail in such situations, primarily because they rely on smoothness and continuity of the underlying manifold, which is violated in the scenarios explored in this paper. However, the proposed algorithm is able to learn effectively in presence of multiple, and potentially intersecting, manifolds, while allowing for the input data to arrive as a massive stream.

stat.ML↗

arXiv2017

Server, server in the cloud. Who is the fairest in the crowd?

This paper follows the recent history of automated beauty competitions to discuss how machine learning techniques, in particular neural networks, alter the way attractiveness is handled and how this impacts the cultural landscape. We describe experiments performed to probe the behavior of two different convolutional neural network architectures in the classification of facial attractiveness in a large database of celebrity faces. As opposed to other easily definable facial features, attractiveness is difficult to detect robustly even for the best classification systems. Based on the observations from these experiments, we discuss several approaches to detect factors that come into play when a machine evaluates human features, and how bias can occur not only in data selection but in network architectures; in multiple forms on multiple levels throughout the process. The overall goal is to map out with mixed methods a novel condition: slippages produced by platform level machine learning systems that make judgements in domains considered dependent on high level human intelligence.

cs.CY↗

arXiv2017

Modeling Graphs Using a Mixture of Kronecker Models

Generative models for graphs are increasingly becoming a popular tool for researchers to generate realistic approximations of graphs. While in the past, focus was on generating graphs which follow general laws, such as the power law for degree distribution, current models have the ability to learn from observed graphs and generate synthetic approximations. The primary emphasis of existing models has been to closely match different properties of a single observed graph. Such models, though stochastic, tend to generate samples which do not have significant variance in terms of the various graph properties. We argue that in many cases real graphs are sampled drawn from a graph population (e.g., networks sampled at various time points, social networks for individual schools, healthcare networks for different geographic regions, etc.). Such populations typically exhibit significant variance. However, existing models are not designed to model this variance, which could lead to issues such as overfitting. We propose a graph generative model that focuses on matching the properties of real graphs and the natural variance expected for the corresponding population. The proposed model adopts a mixture-model strategy to expand the expressiveness of Kronecker product based graph models (KPGM), while building upon the two strengths of KPGM, viz., ability to model several key properties of graphs and to scale to massive graph sizes using its elegant fractal growth based formulation. The proposed model, called x-Kronecker Product Graph Model, or xKPGM, allows scalable learning from observed graphs and generates samples that match the mean and variance of several salient graph properties. We experimentally demonstrate the capability of the proposed model to capture the inherent variability in real world graphs on a variety of publicly available graph data sets.

cs.SI↗

arXiv2017

Tracking System Behaviour from Resource Usage Data

Resource usage data, collected using tools such as TACC Stats, capture the resource utilization by nodes within a high performance computing system. We present methods to analyze the resource usage data to understand the system performance and identify performance anomalies. The core idea is to model the data as a three-way tensor corresponding to the compute nodes, usage metrics, and time. Using the reconstruction error between the original tensor and the tensor reconstructed from a low rank tensor decomposition, as a scalar performance metric, enables us to monitor the performance of the system in an online fashion. This error statistic is then used for anomaly detection that relies on the assumption that the normal/routine behavior of the system can be captured using a low rank approx- imation of the original tensor. We evaluate the performance of the algorithm using information gathered from system logs and show that the performance anomalies identified by the proposed method correlates with critical errors reported in the system logs. Results are shown for data collected for 2013 from the Lonestar4 system at the Texas Advanced Computing Center (TACC)

cs.PF↗

arXiv2017

Error Metrics for Learning Reliable Manifolds from Streaming Data

Spectral dimensionality reduction is frequently used to identify low-dimensional structure in high-dimensional data. However, learning manifolds, especially from the streaming data, is computationally and memory expensive. In this paper, we argue that a stable manifold can be learned using only a fraction of the stream, and the remaining stream can be mapped to the manifold in a significantly less costly manner. Identifying the transition point at which the manifold is stable is the key step. We present error metrics that allow us to identify the transition point for a given stream by quantitatively assessing the quality of a manifold learned using Isomap. We further propose an efficient mapping algorithm, called S-Isomap, that can be used to map new samples onto the stable manifold. We describe experiments on a variety of data sets that show that the proposed approach is computationally efficient without sacrificing accuracy.

stat.ML↗

arXiv2016

Summarizing Large Query Logs in Ettu

Database access logs are large, unwieldy, and hard for humans to inspect and summarize. In spite of this, they remain the canonical go-to resource for tasks ranging from performance tuning to security auditing. In this paper, we address the challenge of compactly encoding large sequences of SQL queries for presentation to a human user. Our approach is based on the Weisfeiler-Lehman (WL) approximate graph isomorphism algorithm, which identifies salient features of a graph or in our case of an abstract syntax tree. Our generalization of WL allows us to define a distance metric for SQL queries, which in turn permits automated clustering of queries. We also present two techniques for visualizing query clusters, and an algorithm that allows these visualizations to be constructed at interactive speeds. Finally, we evaluate our algorithms in the context of a motivating example: insider threat detection at a large US bank. We show experimentally on real world query logs that (a) our distance metric captures a meaningful notion of similarity, and (b) the log summarization process is scalable and performant.

cs.DB↗