arXiv subjects

Naren Ramakrishnan

Publications and source records attributed to Naren Ramakrishnan.

At least 109 records · Page 6Linked to original sources

arXiv2016

Flow of Information in Feed-Forward Deep Neural Networks

Feed-forward deep neural networks have been used extensively in various machine learning applications. Developing a precise understanding of the underling behavior of neural networks is crucial for their efficient deployment. In this paper, we use an information theoretic approach to study the flow of information in a neural network and to determine how entropy of information changes between consecutive layers. Moreover, using the Information Bottleneck principle, we develop a constrained optimization problem that can be used in the training process of a deep neural network. Furthermore, we determine a lower bound for the level of data representation that can be achieved in a deep neural network with an acceptable level of distortion.

cs.IT↗

arXiv2016

Forecasting the Flu: Designing Social Network Sensors for Epidemics

Early detection and modeling of a contagious epidemic can provide important guidance about quelling the contagion, controlling its spread, or the effective design of countermeasures. A topic of recent interest has been to design social network sensors, i.e., identifying a small set of people who can be monitored to provide insight into the emergence of an epidemic in a larger population. We formally pose the problem of designing social network sensors for flu epidemics and identify two different objectives that could be targeted in such sensor design problems. Using the graph theoretic notion of dominators we develop an efficient and effective heuristic for forecasting epidemics at lead time. Using six city-scale datasets generated by extensive microscopic epidemiological simulations involving millions of individuals, we illustrate the practical applicability of our methods and show significant benefits (up to twenty-two days more lead time) compared to other competitors. Most importantly, we demonstrate the use of surrogates or proxies for policy makers for designing social network sensors that require from nonintrusive knowledge of people to more information on the relationship among people. The results show that the more intrusive information we obtain, the longer lead time to predict the flu outbreak up to nine days.

cs.SI↗

arXiv2016

Generating Realistic Synthetic Population Datasets

Modern studies of societal phenomena rely on the availability of large datasets capturing attributes and activities of synthetic, city-level, populations. For instance, in epidemiology, synthetic population datasets are necessary to study disease propagation and intervention measures before implementation. In social science, synthetic population datasets are needed to understand how policy decisions might affect preferences and behaviors of individuals. In public health, synthetic population datasets are necessary to capture diagnostic and procedural characteristics of patient records without violating confidentialities of individuals. To generate such datasets over a large set of categorical variables, we propose the use of the maximum entropy principle to formalize a generative model such that in a statistically well-founded way we can optimally utilize given prior information about the data, and are unbiased otherwise. An efficient inference algorithm is designed to estimate the maximum entropy model, and we demonstrate how our approach is adept at estimating underlying data distributions. We evaluate this approach against both simulated data and on US census datasets, and demonstrate its feasibility using an epidemic simulation application.

cs.DB↗

arXiv2016

Interactive Storytelling over Document Collections

Storytelling algorithms aim to 'connect the dots' between disparate documents by linking starting and ending documents through a series of intermediate documents. Existing storytelling algorithms are based on notions of coherence and connectivity, and thus the primary way by which users can steer the story construction is via design of suitable similarity functions. We present an alternative approach to storytelling wherein the user can interactively and iteratively provide 'must use' constraints to preferentially support the construction of some stories over others. The three innovations in our approach are distance measures based on (inferred) topic distributions, the use of constraints to define sets of linear inequalities over paths, and the introduction of slack and surplus variables to condition the topic distribution to preferentially emphasize desired terms over others. We describe experimental results to illustrate the effectiveness of our interactive storytelling approach over multiple text datasets.

cs.AI↗

arXiv2015

Interactive Discovery of Coordinated Relationship Chains with Maximum Entropy Models

Modern visual analytic tools promote human-in-the-loop analysis but are limited in their ability to direct the user toward interesting and promising directions of study. This problem is especially acute when the analysis task is exploratory in nature, e.g., the discovery of potentially coordinated relationships in massive text datasets. Such tasks are very common in domains like intelligence analysis and security forensics where the goal is to uncover surprising coalitions bridging multiple types of relations. We introduce new maximum entropy models to discover surprising chains of relationships leveraging count data about entity occurrences in documents. These models are embedded in a visual analytic system called MERCER that treats relationship bundles as first class objects and directs the user toward promising lines of inquiry. We demonstrate how user input can judiciously direct analysis toward valid conclusions whereas a purely algorithmic approach could be led astray. Experimental results on both synthetic and real datasets from the intelligence community are presented.

cs.DB↗

arXiv2014

'Beating the news' with EMBERS: Forecasting Civil Unrest using Open Source Indicators

We describe the design, implementation, and evaluation of EMBERS, an automated, 24x7 continuous system for forecasting civil unrest across 10 countries of Latin America using open source indicators such as tweets, news sources, blogs, economic indicators, and other data sources. Unlike retrospective studies, EMBERS has been making forecasts into the future since Nov 2012 which have been (and continue to be) evaluated by an independent T&E team (MITRE). Of note, EMBERS has successfully forecast the uptick and downtick of incidents during the June 2013 protests in Brazil. We outline the system architecture of EMBERS, individual models that leverage specific data sources, and a fusion and suppression engine that supports trading off specific evaluation criteria. EMBERS also provides an audit trail interface that enables the investigation of why specific predictions were made along with the data utilized for forecasting. Through numerous evaluations, we demonstrate the superiority of EMBERS over baserate methods and its capability to forecast significant societal happenings.

cs.SI↗

arXiv2012

Streaming Algorithms for Pattern Discovery over Dynamically Changing Event Sequences

Discovering frequent episodes over event sequences is an important data mining task. In many applications, events constituting the data sequence arrive as a stream, at furious rates, and recent trends (or frequent episodes) can change and drift due to the dynamical nature of the underlying event generation process. The ability to detect and track such the changing sets of frequent episodes can be valuable in many application scenarios. Current methods for frequent episode discovery are typically multipass algorithms, making them unsuitable in the streaming context. In this paper, we propose a new streaming algorithm for discovering frequent episodes over a window of recent events in the stream. Our algorithm processes events as they arrive, one batch at a time, while discovering the top frequent episodes over a window consisting of several batches in the immediate past. We derive approximation guarantees for our algorithm under the condition that frequent episodes are approximately well-separated from infrequent ones in every batch of the window. We present extensive experimental evaluations of our algorithm on both real and synthetic data. We also present comparisons with baselines and adaptations of streaming algorithms from itemset mining literature.

cs.LG↗

arXiv2010

Efficiently Discovering Hammock Paths from Induced Similarity Networks

Similarity networks are important abstractions in many information management applications such as recommender systems, corpora analysis, and medical informatics. For instance, by inducing similarity networks between movies rated similarly by users, or between documents containing common terms, and or between clinical trials involving the same themes, we can aim to find the global structure of connectivities underlying the data, and use the network as a basis to make connections between seemingly disparate entities. In the above applications, composing similarities between objects of interest finds uses in serendipitous recommendation, in storytelling, and in clinical diagnosis, respectively. We present an algorithmic framework for traversing similarity paths using the notion of `hammock' paths which are generalization of traditional paths. Our framework is exploratory in nature so that, given starting and ending objects of interest, it explores candidate objects for path following, and heuristics to admissibly estimate the potential for paths to lead to a desired destination. We present three diverse applications: exploring movie similarities in the Netflix dataset, exploring abstract similarities across the PubMed corpus, and exploring description similarities in a database of clinical trials. Experimental results demonstrate the potential of our approach for unstructured knowledge discovery in similarity networks.

cs.AI↗

arXiv2009

Recommender Systems for the Conference Paper Assignment Problem

Conference paper assignment, i.e., the task of assigning paper submissions to reviewers, presents multi-faceted issues for recommender systems research. Besides the traditional goal of predicting `who likes what?', a conference management system must take into account aspects such as: reviewer capacity constraints, adequate numbers of reviews for papers, expertise modeling, conflicts of interest, and an overall distribution of assignments that balances reviewer preferences with conference objectives. Among these, issues of modeling preferences and tastes in reviewing have traditionally been studied separately from the optimization of paper-reviewer assignment. In this paper, we present an integrated study of both these aspects. First, due to the paucity of data per reviewer or per paper (relative to other recommender systems applications) we show how we can integrate multiple sources of information to learn paper-reviewer preference models. Second, our models are evaluated not just in terms of prediction accuracy but in terms of the end-assignment quality. Using a linear programming-based assignment optimization formulation, we show how our approach better explores the space of unsupplied assignments to maximize the overall affinities of papers assigned to reviewers. We demonstrate our results on real reviewer preference data from the IEEE ICDM 2007 conference.

cs.IR↗

arXiv2009

Towards Chip-on-Chip Neuroscience: Fast Mining of Frequent Episodes Using Graphics Processors

Computational neuroscience is being revolutionized with the advent of multi-electrode arrays that provide real-time, dynamic, perspectives into brain function. Mining event streams from these chips is critical to understanding the firing patterns of neurons and to gaining insight into the underlying cellular activity. We present a GPGPU solution to mining spike trains. We focus on mining frequent episodes which captures coordinated events across time even in the presence of intervening background/"junk" events. Our algorithmic contributions are two-fold: MapConcatenate, a new computation-to-core mapping scheme, and a two-pass elimination approach to quickly find supported episodes from a large number of candidates. Together, they help realize a real-time "chip-on-chip" solution to neuroscience data mining, where one chip (the multi-electrode array) supplies the spike train data and another (the GPGPU) mines it at a scale unachievable previously. Evaluation on both synthetic and real datasets demonstrate the potential of our approach.

cs.DC↗

arXiv2009

Accelerator-Oriented Algorithm Transformation for Temporal Data Mining

Temporal data mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train data. While application scientists have been able to readily gather multi-neuronal datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia's GTX 280 offers a cost-effective option to bring these capabilities to the neuroscientist's desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting "in-the-large" issues such as problem decomposition as well as "in-the-small" issues such as data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly data-parallel mapping strategies. Applications to many datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.

cs.DC↗

arXiv2009

Inferring Dynamic Bayesian Networks using Frequent Episode Mining

Motivation: Several different threads of research have been proposed for modeling and mining temporal data. On the one hand, approaches such as dynamic Bayesian networks (DBNs) provide a formal probabilistic basis to model relationships between time-indexed random variables but these models are intractable to learn in the general case. On the other, algorithms such as frequent episode mining are scalable to large datasets but do not exhibit the rigorous probabilistic interpretations that are the mainstay of the graphical models literature. Results: We present a unification of these two seemingly diverse threads of research, by demonstrating how dynamic (discrete) Bayesian networks can be inferred from the results of frequent episode mining. This helps bridge the modeling emphasis of the former with the counting emphasis of the latter. First, we show how, under reasonable assumptions on data characteristics and on influences of random variables, the optimal DBN structure can be computed using a greedy, local, algorithm. Next, we connect the optimality of the DBN structure with the notion of fixed-delay episodes and their counts of distinct occurrences. Finally, to demonstrate the practical feasibility of our approach, we focus on a specific (but broadly applicable) class of networks, called excitatory networks, and show how the search for the optimal DBN structure can be conducted using just information from frequent episodes. Application on datasets gathered from mathematical models of spiking neurons as well as real neuroscience datasets are presented. Availability: Algorithmic implementations, simulator codebases, and datasets are available from our website at http://neural-code.cs.vt.edu/dbn

cs.LG↗

arXiv2005

A Model Based Approach to Reachability Routing

Current directions in network routing research have not kept pace with the latest developments in network architectures, such as peer-to-peer networks, sensor networks, ad-hoc wireless networks, and overlay networks. A common characteristic among all of these new technologies is the presence of highly dynamic network topologies. Currently deployed single-path routing protocols cannot adequately cope with this dynamism, and existing multi-path algorithms make trade-offs which lead to less than optimal performance on these networks. This drives the need for routing protocols designed with the unique characteristics of these networks in mind. In this paper we propose the notion of reachability routing as a solution to the challenges posed by routing on such dynamic networks. In particular, our formulation of reachability routing provides cost-sensitive multi-path forwarding along with loop avoidance within the confines of the Internet Protocol (IP) architecture. This is achieved through the application of reinforcement learning within a probabilistic routing framework. Following an explanation of our design decisions and a description of the algorithm, we provide an evaluation of the performance of the algorithm on a variety of network topologies. The results show consistently superior performance compared to other reinforcement learning based routing algorithms.

cs.NI↗

arXiv2004

Automatically Generating Interfaces for Personalized Interaction with Digital Libraries

We present an approach to automatically generate interfaces supporting personalized interaction with digital libraries; these interfaces augment the user-DL dialog by empowering the user to (optionally) supply out-of-turn information during an interaction, flatten or restructure the dialog, and enquire about dialog options. Interfaces generated using this approach for CITIDEL are described.

cs.DL↗

arXiv2003

Taking the Initiative with Extempore: Exploring Out-of-Turn Interactions with Websites

We present the first study to explore the use of out-of-turn interaction in websites. Out-of-turn interaction is a technique which empowers the user to supply unsolicited information while browsing. This approach helps flexibly bridge any mental mismatch between the user and the website, in a manner fundamentally different from faceted browsing and site-specific search tools. We built a user interface (Extempore) which accepts out-of-turn input via voice or text; and employed it in a US congressional website, to determine if users utilize out-of-turn interaction for information-finding tasks, and their rationale for doing so. The results indicate that users are adept at discerning when out-of-turn interaction is necessary in a particular task, and actively interleaved it with browsing. However, users found cascading information across information-finding subtasks challenging. Therefore, this work not only improves our understanding of out-of-turn interaction, but also suggests further opportunities to enrich browsing experiences for users.

cs.HC↗

arXiv2003

Turning CARTwheels: An Alternating Algorithm for Mining Redescriptions

We present an unusual algorithm involving classification trees where two trees are grown in opposite directions so that they are matched at their leaves. This approach finds application in a new data mining task we formulate, called "redescription mining". A redescription is a shift-of-vocabulary, or a different way of communicating information about a given subset of data; the goal of redescription mining is to find subsets of data that afford multiple descriptions. We highlight the importance of this problem in domains such as bioinformatics, which exhibit an underlying richness and diversity of data descriptors (e.g., genes can be studied in a variety of ways). Our approach helps integrate multiple forms of characterizing datasets, situates the knowledge gained from one dataset in the context of others, and harnesses high-level abstractions for uncovering cryptic and subtle features of data. Algorithm design decisions, implementation details, and experimental results are presented.

cs.CE↗

arXiv2003

Staging Transformations for Multimodal Web Interaction Management

Multimodal interfaces are becoming increasingly ubiquitous with the advent of mobile devices, accessibility considerations, and novel software technologies that combine diverse interaction media. In addition to improving access and delivery capabilities, such interfaces enable flexible and personalized dialogs with websites, much like a conversation between humans. In this paper, we present a software framework for multimodal web interaction management that supports mixed-initiative dialogs between users and websites. A mixed-initiative dialog is one where the user and the website take turns changing the flow of interaction. The framework supports the functional specification and realization of such dialogs using staging transformations -- a theory for representing and reasoning about dialogs based on partial input. It supports multiple interaction interfaces, and offers sessioning, caching, and co-ordination functions through the use of an interaction manager. Two case studies are presented to illustrate the promise of this approach.

cs.IR↗

arXiv2003

Supporting Out-of-turn Interactions in a Multimodal Web Interface

Multimodal interfaces are becoming increasingly important with the advent of mobile devices, accessibility considerations, and novel software technologies that combine diverse interaction media. This article investigates systems support for web browsing in a multimodal interface. Specifically, we outline the design and implementation of a software framework that integrates hyperlink and speech modes of interaction. Instead of viewing speech as merely an alternative interaction medium, the framework uses it to support out-of-turn interaction, providing a flexibility of information access not possible with hyperlinks alone. This approach enables the creation of websites that adapt to the needs of users, yet permits the designer fine-grained control over what interactions to support. Design methodology, implementation details, and two case studies are presented.

cs.IR↗