Matching Items (171)
Description
The rapid escalation of technology and the widespread emergence of modern technological equipments have resulted in the generation of humongous amounts of digital data (in the form of images, videos and text). This has expanded the possibility of solving real world problems using computational learning frameworks. However, while gathering a large amount of data is cheap and easy, annotating them with class labels is an expensive process in terms of time, labor and human expertise. This has paved the way for research in the field of active learning. Such algorithms automatically select the salient and exemplar instances from large quantities of unlabeled data and are effective in reducing human labeling effort in inducing classification models. To utilize the possible presence of multiple labeling agents, there have been attempts towards a batch mode form of active learning, where a batch of data instances is selected simultaneously for manual annotation. This dissertation is aimed at the development of novel batch mode active learning algorithms to reduce manual effort in training classification models in real world multimedia pattern recognition applications. Four major contributions are proposed in this work: $(i)$ a framework for dynamic batch mode active learning, where the batch size and the specific data instances to be queried are selected adaptively through a single formulation, based on the complexity of the data stream in question, $(ii)$ a batch mode active learning strategy for fuzzy label classification problems, where there is an inherent imprecision and vagueness in the class label definitions, $(iii)$ batch mode active learning algorithms based on convex relaxations of an NP-hard integer quadratic programming (IQP) problem, with guaranteed bounds on the solution quality and $(iv)$ an active matrix completion algorithm and its application to solve several variants of the active learning problem (transductive active learning, multi-label active learning, active feature acquisition and active learning for regression). These contributions are validated on the face recognition and facial expression recognition problems (which are commonly encountered in real world applications like robotics, security and assistive technology for the blind and the visually impaired) and also on collaborative filtering applications like movie recommendation.
ContributorsChakraborty, Shayok (Author) / Panchanathan, Sethuraman (Thesis advisor) / Balasubramanian, Vineeth N. (Committee member) / Li, Baoxin (Committee member) / Mittelmann, Hans (Committee member) / Ye, Jieping (Committee member) / Arizona State University (Publisher)
Created2013
Description
The increasing popularity of Twitter renders improved trustworthiness and relevance assessment of tweets much more important for search. However, given the limitations on the size of tweets, it is hard to extract measures for ranking from the tweet's content alone. I propose a method of ranking tweets by generating a reputation score for each tweet that is based not just on content, but also additional information from the Twitter ecosystem that consists of users, tweets, and the web pages that tweets link to. This information is obtained by modeling the Twitter ecosystem as a three-layer graph. The reputation score is used to power two novel methods of ranking tweets by propagating the reputation over an agreement graph based on tweets' content similarity. Additionally, I show how the agreement graph helps counter tweet spam. An evaluation of my method on 16~million tweets from the TREC 2011 Microblog Dataset shows that it doubles the precision over baseline Twitter Search and achieves higher precision than current state of the art method. I present a detailed internal empirical evaluation of RAProp in comparison to several alternative approaches proposed by me, as well as external evaluation in comparison to the current state of the art method.
ContributorsRavikumar, Srijith (Author) / Kambhampati, Subbarao (Thesis advisor) / Davulcu, Hasan (Committee member) / Liu, Huan (Committee member) / Arizona State University (Publisher)
Created2013
Description
Continuous Delivery, as one of the youngest and most popular member of agile model family, has become a popular concept and method in software development industry recently. Instead of the traditional software development method, which requirements and solutions must be fixed before starting software developing, it promotes adaptive planning, evolutionary development and delivery, and encourages rapid and flexible response to change. However, several problems prevent Continuous Delivery to be introduced into education world. Taking into the consideration of the barriers, we propose a new Cloud based Continuous Delivery Software Developing System. This system is designed to fully utilize the whole life circle of software developing according to Continuous Delivery concepts in a virtualized environment in Vlab platform.
ContributorsDeng, Yuli (Author) / Huang, Dijiang (Thesis advisor) / Davulcu, Hasan (Committee member) / Chen, Yinong (Committee member) / Arizona State University (Publisher)
Created2013
Description
Our research focuses on finding answers through decentralized search, for complex, imprecise queries (such as "Which is the best hair salon nearby?") in situations where there is a spatiotemporal constraint (say answer needs to be found within 15 minutes) associated with the query. In general, human networks are good in answering imprecise queries. We try to use the social network of a person to answer his query. Our research aims at designing a framework that exploits the user's social network in order to maximize the answers for a given query. Exploiting an user's social network has several challenges. The major challenge is that the user's immediate social circle may not possess the answer for the given query, and hence the framework designed needs to carry out the query diffusion process across the network. The next challenge involves in finding the right set of seeds to pass the query to in the user's social circle. One other challenge is to incentivize people in the social network to respond to the query and thereby maximize the quality and quantity of replies. Our proposed framework is a mobile application where an individual can either respond to the query or forward it to his friends. We simulated the query diffusion process in three types of graphs: Small World, Random and Preferential Attachment. Given a type of network and a particular query, we carried out the query diffusion by selecting seeds based on attributes of the seed. The main attributes are Topic relevance, Replying or Forwarding probability and Time to Respond. We found that there is a considerable increase in the number of replies attained, even without saturating the user's network, if we adopt an optimal seed selection process. We found the output of the optimal algorithm to be satisfactory as the number of replies received at the interrogator's end was close to three times the number of neighbors an interrogator has. We addressed the challenge of incentivizing people to respond by associating a particular amount of points for each query asked, and awarding the same to people involved in answering the query. Thus, we aim to design a mobile application based on our proposed framework so that it helps in maximizing the replies for the interrogator's query by diffusing the query across his/her social network.
ContributorsSwaminathan, Neelakantan (Author) / Sundaram, Hari (Thesis advisor) / Davulcu, Hasan (Thesis advisor) / Turaga, Pavan (Committee member) / Arizona State University (Publisher)
Created2013
Description
With the advent of social media (like Twitter, Facebook etc.,) people are easily sharing their opinions, sentiments and enforcing their ideologies on others like never before. Even people who are otherwise socially inactive would like to share their thoughts on current affairs by tweeting and sharing news feeds with their friends and acquaintances. In this thesis study, we chose Twitter as our main data platform to analyze shifts and movements of 27 political organizations in Indonesia. So far, we have collected over 30 million tweets and 150,000 news articles from RSS feeds of the corresponding organizations for our analysis. For Twitter data extraction, we developed a multi-threaded application which seamlessly extracts, cleans and stores millions of tweets matching our keywords from Twitter Streaming API. For keyword extraction, we used topics and perspectives which were extracted using n-grams techniques and later approved by our social scientists. After the data is extracted, we aggregate the tweet contents that belong to every user on a weekly basis. Finally, we applied linear and logistic regression using SLEP, an open source sparse learning package to compute weekly score for users and mapping them to one of the 27 organizations on a radical or counter radical scale. Since, we are mapping users to organizations on a weekly basis, we are able to track user's behavior and important new events that triggered shifts among users between organizations. This thesis study can further be extended to identify topics and organization specific influential users and new users from various social media platforms like Facebook, YouTube etc. can easily be mapped to existing organizations on a radical or counter-radical scale.
ContributorsPoornachandran, Sathishkumar (Author) / Davulcu, Hasan (Thesis advisor) / Sen, Arunabha (Committee member) / Woodward, Mark (Committee member) / Arizona State University (Publisher)
Created2013
Description
In recent years, machine learning and data mining technologies have received growing attention in several areas such as recommendation systems, natural language processing, speech and handwriting recognition, image processing and biomedical domain. Many of these applications which deal with physiological and biomedical data require person specific or person adaptive systems. The greatest challenge in developing such systems is the subject-dependent data variations or subject-based variability in physiological and biomedical data, which leads to difference in data distributions making the task of modeling these data, using traditional machine learning algorithms, complex and challenging. As a result, despite the wide application of machine learning, efficient deployment of its principles to model real-world data is still a challenge. This dissertation addresses the problem of subject based variability in physiological and biomedical data and proposes person adaptive prediction models based on novel transfer and active learning algorithms, an emerging field in machine learning. One of the significant contributions of this dissertation is a person adaptive method, for early detection of muscle fatigue using Surface Electromyogram signals, based on a new multi-source transfer learning algorithm. This dissertation also proposes a subject-independent algorithm for grading the progression of muscle fatigue from 0 to 1 level in a test subject, during isometric or dynamic contractions, at real-time. Besides subject based variability, biomedical image data also varies due to variations in their imaging techniques, leading to distribution differences between the image databases. Hence a classifier learned on one database may perform poorly on the other database. Another significant contribution of this dissertation has been the design and development of an efficient biomedical image data annotation framework, based on a novel combination of transfer learning and a new batch-mode active learning method, capable of addressing the distribution differences across databases. The methodologies developed in this dissertation are relevant and applicable to a large set of computing problems where there is a high variation of data between subjects or sources, such as face detection, pose detection and speech recognition. From a broader perspective, these frameworks can be viewed as a first step towards design of automated adaptive systems for real world data.
ContributorsChattopadhyay, Rita (Author) / Panchanathan, Sethuraman (Thesis advisor) / Ye, Jieping (Thesis advisor) / Li, Baoxin (Committee member) / Santello, Marco (Committee member) / Arizona State University (Publisher)
Created2013
Description
This dissertation presents the Temporal Event Query Language (TEQL), a new language for querying event streams. Event Stream Processing enables online querying of streams of events to extract relevant data in a timely manner. TEQL enables querying of interval-based event streams using temporal database operators. Temporal databases and temporal query languages have been a subject of research for more than 30 years and are a natural fit for expressing queries that involve a temporal dimension. However, operators developed in this context cannot be directly applied to event streams. The research extends a preexisting relational framework for event stream processing to support temporal queries. The language features and formal semantic extensions to extend the relational framework are identified. The extended framework supports continuous, step-wise evaluation of temporal queries. The incremental evaluation of TEQL operators is formalized to avoid re-computation of previous results. The research includes the development of a prototype that supports the integrated event and temporal query processing framework, with support for incremental evaluation and materialization of intermediate results. TEQL enables reporting temporal data in the output, direct specification of conditions over timestamps, and specification of temporal relational operators. Through the integration of temporal database operators with event languages, a new class of temporal queries is made possible for querying event streams. New features include semantic aggregation, extraction of temporal patterns using set operators, and a more accurate specification of event co-occurrence.
ContributorsShiva, Foruhar Ali (Author) / Urban, Susan D (Thesis advisor) / Chen, Yi (Thesis advisor) / Davulcu, Hasan (Committee member) / Sarjoughian, Hessam S. (Committee member) / Arizona State University (Publisher)
Created2012
Description
The pay-as-you-go economic model of cloud computing increases the visibility, traceability, and verifiability of software costs. Application developers must understand how their software uses resources when running in the cloud in order to stay within budgeted costs and/or produce expected profits. Cloud computing's unique economic model also leads naturally to an earn-as-you-go profit model for many cloud based applications. These applications can benefit from low level analyses for cost optimization and verification. Testing cloud applications to ensure they meet monetary cost objectives has not been well explored in the current literature. When considering revenues and costs for cloud applications, the resource economic model can be scaled down to the transaction level in order to associate source code with costs incurred while running in the cloud. Both static and dynamic analysis techniques can be developed and applied to understand how and where cloud applications incur costs. Such analyses can help optimize (i.e. minimize) costs and verify that they stay within expected tolerances. An adaptation of Worst Case Execution Time (WCET) analysis is presented here to statically determine worst case monetary costs of cloud applications. This analysis is used to produce an algorithm for determining control flow paths within an application that can exceed a given cost threshold. The corresponding results are used to identify path sections that contribute most to cost excess. A hybrid approach for determining cost excesses is also presented that is comprised mostly of dynamic measurements but that also incorporates calculations that are based on the static analysis approach. This approach uses operational profiles to increase the precision and usefulness of the calculations.
ContributorsBuell, Kevin, Ph.D (Author) / Collofello, James (Thesis advisor) / Davulcu, Hasan (Committee member) / Lindquist, Timothy (Committee member) / Sen, Arunabha (Committee member) / Arizona State University (Publisher)
Created2012
Description
A semiconductor supply chain modeling and simulation platform using Linear Program (LP) optimization and parallel Discrete Event System Specification (DEVS) process models has been developed in a joint effort by ASU and Intel Corporation. A Knowledge Interchange Broker (KIBDEVS/LP) was developed to broker information synchronously between the DEVS and LP models. Recently a single-echelon heuristic Inventory Strategy Module (ISM) was added to correct for forecast bias in customer demand data using different smoothing techniques. The optimization model could then use information provided by the forecast model to make better decisions for the process model. The composition of ISM with LP and DEVS models resulted in the first realization of what is now called the Optimization Simulation Forecast (OSF) platform. It could handle a single echelon supply chain system consisting of single hubs and single products In this thesis, this single-echelon simulation platform is extended to handle multiple echelons with multiple inventory elements handling multiple products. The main aspect for the multi-echelon OSF platform was to extend the KIBDEVS/LP such that ISM interactions with the LP and DEVS models could also be supported. To achieve this, a new, scalable XML schema for the KIB has been developed. The XML schema has also resulted in strengthening the KIB execution engine design. A sequential scheme controls the executions of the DEVS-Suite simulator, CPLEX optimizer, and ISM engine. To use the ISM for multiple echelons, it is extended to compute forecast customer demands and safety stocks over multiple hubs and products. Basic examples for semiconductor manufacturing spanning single and two echelon supply chain systems have been developed and analyzed. Experiments using perfect data were conducted to show the correctness of the OSF platform design and implementation. Simple, but realistic experiments have also been conducted. They highlight the kinds of supply chain dynamics that can be evaluated using discrete event process simulation, linear programming optimization, and heuristics forecasting models.
ContributorsSmith, James Melkon (Author) / Sarjoughian, Hessam S. (Thesis advisor) / Davulcu, Hasan (Committee member) / Fainekos, Georgios (Committee member) / Arizona State University (Publisher)
Created2012
Description
Process migration is a heavily studied research area and has a number of applications in distributed systems. Process migration means transferring a process running on one machine to another such that it resumes execution from the point at which it was suspended. The conventional approach to implement process migration is to move the entire state information of the process (including hardware context, virtual memory, files etc.) from one machine to another. Copying all the state information is costly. This thesis proposes and demonstrates a new approach of migrating a process between two cores of Intel Single Chip Cloud (SCC), an experimental 48-core processor by Intel, with each core running a separate instance of the operating system. In this method the amount of process state to be transferred from one core's memory to another is reduced by making use of special registers called Lookup tables (LUTs) present on each core of SCC. Thus this new approach is faster than the conventional method.
ContributorsJain, Vaibhav (Author) / Dasgupta, Partha (Thesis advisor) / Shriavstava, Aviral (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2013