Open Access. Powered by Scholars. Published by Universities.®
- Keyword
-
- Apache Spark (1)
- Best Trade-off Point (1)
- Control Architecture (1)
- Control Systems Engineering (1)
- Crowdsourcing (1)
-
- Deep Learning (1)
- Desktop (1)
- Dynamic resource allocation (1)
- Elbow curve (1)
- Energy efficiency (1)
- Exponential decay curve (1)
- Eyetracking (1)
- Hadoop MapReduce (1)
- Human attention region of interest (1)
- Human–Computer Information Retrieval (1)
- Just noticeable difference (1)
- Machine Learning (1)
- Mobile (1)
- Multilingual Interfaces (1)
- Multilingual Search (1)
- Optimal resource provisioning (1)
- Perceptual video coding (1)
- Performance efficiency (1)
- Recommender Systems (1)
- Saturation (1)
- Scalar Field Adaptive Navigation (1)
- Swarm Aggregation (1)
- Swarm Robotics (1)
- User Studies (1)
- Video Coding (1)
Articles 1 - 14 of 14
Full-Text Articles in Engineering
On Sparse Coding As An Alternate Transform In Video Coding, Michael G. Schimpf
On Sparse Coding As An Alternate Transform In Video Coding, Michael G. Schimpf
Engineering Ph.D. Theses
In video compression, specifically in the prediction process, a residual signal is calculated by subtracting the predicted from the original signal, which represents the error of this process. This residual signal is usually transformed by a discrete cosine transform (DCT) from the pixel, into the frequency domain. It is then quantized, which filters more or less high frequencies (depending on a quality parameter). The quantized signal is then entropy encoded usually by a context-adaptive binary arithmetic coding engine (CABAC), and written into a bitstream. In the decoding phase the process is reversed. DCT and quantization in combination are efficient tools, …
A Submodular Optimization Framework For Imbalanced Text Classification With Data Augmentation, Eyor Alemayehu
A Submodular Optimization Framework For Imbalanced Text Classification With Data Augmentation, Eyor Alemayehu
Engineering Ph.D. Theses
In the domain of text classification, imbalanced datasets are a common occurrence. The skewed distribution of the labels of these datasets poses a great challenge to the performance of text classifiers. One popular way to mitigate this challenge is to augment underwhelmingly represented labels with synthesized items. The synthesized items are generated by data augmentation methods that can typically generate an unbounded number of items. To select the synthesized items that maximize the performance of text classifiers, we introduce a novel method that selects items that jointly maximize the likelihood of the items belonging to their respective labels and the …
Reactive Particle Swarm Control Architecture And Application For Scalar Field Adaptive Navigation, Shae Taylor Hart
Reactive Particle Swarm Control Architecture And Application For Scalar Field Adaptive Navigation, Shae Taylor Hart
Engineering Ph.D. Theses
Adaptive navigation is a subcategory of navigation techniques that attempts to identify goal locations that satisfy specific criteria in an unknown area. In 2D scalar field adaptive navigation (SFAN), primitives navigate to or along features of interest in an unknown, possibly time-varying, planar scalar field. Features include extrema, contours, and fronts. This work solves the 2D SFAN problem using swarm robotic techniques. Robotic swarms are a subset of multi-robot systems that use decentralized control of simple interchangeable robots to perform collective actions. A subgroup of swarms is the Reactive Particle Swarm (RPS), characterized based on its simplicity, reactivity to its …
Metascriptura: A General Data Provenance Framework, Maria Joseph Israel
Metascriptura: A General Data Provenance Framework, Maria Joseph Israel
Engineering Ph.D. Theses
Digital technology makes it easy to generate and distribute large volumes of data. However, it has also complicated the process of verifying and validating sources of data and their derivatives risking obfuscation of truth amidst the deluge of data. To address this issue, I trace and develop an approach based on data provenance tracking. Specifically, I make it possible to deep trace the origins and lineages of data, by applying state-of-the-art data provenance technologies, which I extend beyond traditional data provenance applications. In this dissertation, I demonstrate that with the right data infrastructure it is feasible to grant greater agency …
Enhancing The Quality Of Service And Energy Efficiency Of Wifi-Based Iot Networks, Jaykumar Sheth
Enhancing The Quality Of Service And Energy Efficiency Of Wifi-Based Iot Networks, Jaykumar Sheth
Engineering Ph.D. Theses
The 802.11 standard, known as WiFi, is currently being used for a wide variety of applications including Internet of Things (IoT). However, the contention between the traffic of IoT stations (STAs) as well as the contention between these flows and regular user-generated traffic reduces the energy efficiency and timeliness of IoT communication. To remedy this problem, in this thesis, we take the following approaches for mitigating the challenges faced by WiFi-based IoT networks: First, we highlight the importance of observability with respect to WiFi networks and how it helps the researchers to better examine the dynamics of issues and its …
Personalized Memory Transfer For Conversational Recommendation Systems, Naga Archana Godavarthy
Personalized Memory Transfer For Conversational Recommendation Systems, Naga Archana Godavarthy
Engineering Ph.D. Theses
Dialogue systems are becoming an increasingly common part of many users' daily routines. Natural language serves as a convenient interface to express our preferences with the underlying systems. In this work, we implement a full-fledged Conversational Recommendation System, mainly focusing on learning user preferences through online conversations. Compared to the traditional collaborative filtering setting where feedback is provided quantitatively, conversational users may only indicate their preferences at a high level with inexact item mentions in the form of natural language chit-chat. This makes it harder for the system to correctly interpret user intent and in turn provide useful recommendations to …
Deep Generative Models For Semantic Text Hashing, Suthee Chaidaroon
Deep Generative Models For Semantic Text Hashing, Suthee Chaidaroon
Engineering Ph.D. Theses
As the amount of textual data has been rapidly increasing over the past decade, efficient similarity search methods have become a crucial component of large-scale information retrieval systems. A popular strategy is to represent original data samples by compact binary codes through hashing. A spectrum of machine learning methods have been utilized, but they often lack expressiveness and flexibility in modeling to learn effective representations. The recent advances of deep learning in a wide range of applications has demonstrated its capability to learn robust and powerful feature representations for complex data. Especially, deep generative models naturally combine the expressiveness of …
Mnews: A Study Of Multilingual News Search Interfaces, Chenjun Ling
Mnews: A Study Of Multilingual News Search Interfaces, Chenjun Ling
Engineering Ph.D. Theses
With the global expansion of the Internet and the World Wide Web, users are becoming increasingly diverse, particularly in terms of languages. In fact, the number of polyglot Web users across the globe has increased dramatically.
However, even such multilingual users often continue to suffer from unbalanced and fragmented news information, as traditional news access systems seldom allow users to simultaneously search for and/or compare news in different languages, even though prior research results have shown that multilingual users make significant use of each of their languages when searching for information online.
Relatively little human-centered research has been conducted to …
Deep Learning For Recommender Systems, Travis Akira Ebesu
Deep Learning For Recommender Systems, Travis Akira Ebesu
Engineering Ph.D. Theses
The widespread adoption of the Internet has led to an explosion in the number of choices available to consumers. Users begin to expect personalized content in modern E-commerce, entertainment and social media platforms. Recommender Systems (RS) provide a critical solution to this problem by maintaining user engagement and satisfaction with personalized content.
Traditional RS techniques are often linear limiting the expressivity required to model complex user-item interactions and require extensive handcrafted features from domain experts. Deep learning demonstrated significant breakthroughs in solving problems that have alluded the artificial intelligence community for many years advancing state-of-the-art results in domains such as …
Shingled Magnetic Recording Disks For Mass Storage Systems, Quoc Minh Le
Shingled Magnetic Recording Disks For Mass Storage Systems, Quoc Minh Le
Engineering Ph.D. Theses
Disk drives have seen a dramatic increase in storage density over the last five decades, but to continue the growth seems difficult if not impossible because of physical limitations. One way to increase storage density is using a shingled magnetic recording (SMR) disk. Shingled writing is a promising technique that trades off the inability to update in-place for narrower tracks and thus a much higher data density. It is particularly appealing as it can be adopted while utilizing essentially the same physical recording mechanisms currently in use. Because of its manner of writing, an SMR disk would be unable to …
Human Attention Region Of Interest In Video Compression, Olayinka Sylvia N’Guessan
Human Attention Region Of Interest In Video Compression, Olayinka Sylvia N’Guessan
Engineering Ph.D. Theses
In this thesis, we propose a generic human attention region-of-interest (Generic- HAROI) algorithm to improve video compression while preserving subjective quality. Precisely, this algorithm performs a perceptual adaptive quantization algorithm on video frames as a function of the distribution of their luminance, motion vector, and color saturation. Our research incorporates a psycho-visual study that demonstrated that human attention automatically enhanced perceived saturation. As a result, the adaptive quantization phase of our compression algorithm is characterized by a luminance and saturation-aware just noticeable distortion (JND) function. After running multiple experiments on 18 videos with various resolutions ranging from QCIF to 4K, …
Machine Learning Models For Context-Aware Recommender Systems, Yogesh Jhamb
Machine Learning Models For Context-Aware Recommender Systems, Yogesh Jhamb
Engineering Ph.D. Theses
The mass adoption of the internet has resulted in the exponential growth of products and services on the world wide web. An individual consumer, faced with this data deluge, is expected to make reasonable choices saving time and money. Organizations are facing increased competition, and they are looking for innovative ways to increase revenue and customer loyalty. A business wants to target the right product or service to an individual consumer, and this drives personalized recommendation. Recommender systems, designed to provide personalized recommendations, initially focused only on the user-item interaction. However, these systems evolved to provide a context-aware recommendations. Context-aware …
Towards Efficient Resource Provisioning In Hadoop, Peter P. Nghiem
Towards Efficient Resource Provisioning In Hadoop, Peter P. Nghiem
Engineering Ph.D. Theses
Considering recent exponential growth in the amount of information processed in Big Data, the high energy consumed by data processing engines in datacenters has become a major issue, underlining the need for efficient resource allocation for better energy-efficient computing. This thesis proposes the Best Trade-off Point (BToP) method which provides a general approach and techniques based on an algorithm with mathematical formulas to find the best trade-off point on an elbow curve of performance vs. resources for efficient resource provisioning in Hadoop MapReduce and Apache Spark. Our novel BToP method is expected to work for any applications and systems which …
Supernode Transformation On Parallel Systems With Distributed Memory – An Analytical Approach, Yong Chen
Supernode Transformation On Parallel Systems With Distributed Memory – An Analytical Approach, Yong Chen
Engineering Ph.D. Theses
Supernode transformation, or tiling, is a technique that partitions algorithms to improve data locality and parallelism by balancing computation and inter-processor communication costs to achieve shortest execution or running time. It groups multiple iterations of nested loops into supernodes to be assigned to processors for processing in parallel. A supernode transformation can be described by supernode size and shape. This research focuses on supernode transformation on multi-processor architectures with distributed memory, including computer cluster systems and General Purpose Graphic Processing Units (GPGPUs). The research involves supernode scheduling, supernode mapping to processors, and the finding of the optimal supernode size, for …