Matching Items (3)
Filtering by
- Genre: Doctoral Dissertation
Description
Feature representations for raw data is one of the most important component in a machine learning system. Traditionally, features are \textit{hand crafted} by domain experts which can often be a time consuming process. Furthermore, they do not generalize well to unseen data and novel tasks. Recently, there have been many efforts to generate data-driven representations using clustering and sparse models. This dissertation focuses on building data-driven unsupervised models for analyzing raw data and developing efficient feature representations.
Simultaneous segmentation and feature extraction approaches for silicon-pores sensor data are considered. Aggregating data into a matrix and performing low rank and sparse matrix decompositions with additional smoothness constraints are proposed to solve this problem. Comparison of several variants of the approaches and results for signal de-noising and translocation/trapping event extraction are presented. Algorithms to improve transform-domain features for ion-channel time-series signals based on matrix completion are presented. The improved features achieve better performance in classification tasks and in reducing the false alarm rates when applied to analyte detection.
Developing representations for multimedia is an important and challenging problem with applications ranging from scene recognition, multi-media retrieval and personal life-logging systems to field robot navigation. In this dissertation, we present a new framework for feature extraction for challenging natural environment sounds. Proposed features outperform traditional spectral features on challenging environmental sound datasets. Several algorithms are proposed that perform supervised tasks such as recognition and tag annotation. Ensemble methods are proposed to improve the tag annotation process.
To facilitate the use of large datasets, fast implementations are developed for sparse coding, the key component in our algorithms. Several strategies to speed-up Orthogonal Matching Pursuit algorithm using CUDA kernel on a GPU are proposed. Implementations are also developed for a large scale image retrieval system. Image-based "exact search" and "visually similar search" using the image patch sparse codes are performed. Results demonstrate large speed-up over CPU implementations and good retrieval performance is also achieved.
Simultaneous segmentation and feature extraction approaches for silicon-pores sensor data are considered. Aggregating data into a matrix and performing low rank and sparse matrix decompositions with additional smoothness constraints are proposed to solve this problem. Comparison of several variants of the approaches and results for signal de-noising and translocation/trapping event extraction are presented. Algorithms to improve transform-domain features for ion-channel time-series signals based on matrix completion are presented. The improved features achieve better performance in classification tasks and in reducing the false alarm rates when applied to analyte detection.
Developing representations for multimedia is an important and challenging problem with applications ranging from scene recognition, multi-media retrieval and personal life-logging systems to field robot navigation. In this dissertation, we present a new framework for feature extraction for challenging natural environment sounds. Proposed features outperform traditional spectral features on challenging environmental sound datasets. Several algorithms are proposed that perform supervised tasks such as recognition and tag annotation. Ensemble methods are proposed to improve the tag annotation process.
To facilitate the use of large datasets, fast implementations are developed for sparse coding, the key component in our algorithms. Several strategies to speed-up Orthogonal Matching Pursuit algorithm using CUDA kernel on a GPU are proposed. Implementations are also developed for a large scale image retrieval system. Image-based "exact search" and "visually similar search" using the image patch sparse codes are performed. Results demonstrate large speed-up over CPU implementations and good retrieval performance is also achieved.
ContributorsSattigeri, Prasanna S (Author) / Spanias, Andreas (Thesis advisor) / Thornton, Trevor (Committee member) / Goryll, Michael (Committee member) / Tsakalis, Konstantinos (Committee member) / Arizona State University (Publisher)
Created2014
Description
General-purpose processors propel the advances and innovations that are the subject of humanity’s many endeavors. Catering to this demand, chip-multiprocessors (CMPs) and general-purpose graphics processing units (GPGPUs) have seen many high-performance innovations in their architectures. With these advances, the memory subsystem has become the performance- and energy-limiting aspect of CMPs and GPGPUs alike. This dissertation identifies and mitigates the key performance and energy-efficiency bottlenecks in the memory subsystem of general-purpose processors via novel, practical, microarchitecture and system-architecture solutions.
Addressing the important Last Level Cache (LLC) management problem in CMPs, I observe that LLC management decisions made in isolation, as in prior proposals, often lead to sub-optimal system performance. I demonstrate that in order to maximize system performance, it is essential to manage the LLCs while being cognizant of its interaction with the system main memory. I propose ReMAP, which reduces the net memory access cost by evicting cache lines that either have no reuse, or have low memory access cost. ReMAP improves the performance of the CMP system by as much as 13%, and by an average of 6.5%.
Rather than the LLC, the L1 data cache has a pronounced impact on GPGPU performance by acting as the bandwidth filter for the rest of the memory subsystem. Prior work has shown that the severely constrained data cache capacity in GPGPUs leads to sub-optimal performance. In this thesis, I propose two novel techniques that address the GPGPU data cache capacity problem. I propose ID-Cache that performs effective cache bypassing and cache line size selection to improve cache capacity utilization. Next, I propose LATTE-CC that considers the GPU’s latency tolerance feature and adaptively compresses the data stored in the data cache, thereby increasing its effective capacity. ID-Cache and LATTE-CC are shown to achieve 71% and 19.2% speedup, respectively, over a wide variety of GPGPU applications.
Complementing the aforementioned microarchitecture techniques, I identify the need for system architecture innovations to sustain performance scalability of GPG- PUs in the face of slowing Moore’s Law. I propose a novel GPU architecture called the Multi-Chip-Module GPU (MCM-GPU) that integrates multiple GPU modules to form a single logical GPU. With intelligent memory subsystem optimizations tailored for MCM-GPUs, it can achieve within 7% of the performance of a similar but hypothetical monolithic die GPU. Taking a step further, I present an in-depth study of the energy-efficiency characteristics of future MCM-GPUs. I demonstrate that the inherent non-uniform memory access side-effects form the key energy-efficiency bottleneck in the future.
In summary, this thesis offers key insights into the performance and energy-efficiency bottlenecks in CMPs and GPGPUs, which can guide future architects towards developing high-performance and energy-efficient general-purpose processors.
Addressing the important Last Level Cache (LLC) management problem in CMPs, I observe that LLC management decisions made in isolation, as in prior proposals, often lead to sub-optimal system performance. I demonstrate that in order to maximize system performance, it is essential to manage the LLCs while being cognizant of its interaction with the system main memory. I propose ReMAP, which reduces the net memory access cost by evicting cache lines that either have no reuse, or have low memory access cost. ReMAP improves the performance of the CMP system by as much as 13%, and by an average of 6.5%.
Rather than the LLC, the L1 data cache has a pronounced impact on GPGPU performance by acting as the bandwidth filter for the rest of the memory subsystem. Prior work has shown that the severely constrained data cache capacity in GPGPUs leads to sub-optimal performance. In this thesis, I propose two novel techniques that address the GPGPU data cache capacity problem. I propose ID-Cache that performs effective cache bypassing and cache line size selection to improve cache capacity utilization. Next, I propose LATTE-CC that considers the GPU’s latency tolerance feature and adaptively compresses the data stored in the data cache, thereby increasing its effective capacity. ID-Cache and LATTE-CC are shown to achieve 71% and 19.2% speedup, respectively, over a wide variety of GPGPU applications.
Complementing the aforementioned microarchitecture techniques, I identify the need for system architecture innovations to sustain performance scalability of GPG- PUs in the face of slowing Moore’s Law. I propose a novel GPU architecture called the Multi-Chip-Module GPU (MCM-GPU) that integrates multiple GPU modules to form a single logical GPU. With intelligent memory subsystem optimizations tailored for MCM-GPUs, it can achieve within 7% of the performance of a similar but hypothetical monolithic die GPU. Taking a step further, I present an in-depth study of the energy-efficiency characteristics of future MCM-GPUs. I demonstrate that the inherent non-uniform memory access side-effects form the key energy-efficiency bottleneck in the future.
In summary, this thesis offers key insights into the performance and energy-efficiency bottlenecks in CMPs and GPGPUs, which can guide future architects towards developing high-performance and energy-efficient general-purpose processors.
ContributorsArunkumar, Akhil (Author) / Wu, Carole-Jean (Thesis advisor) / Shrivastava, Aviral (Committee member) / Lee, Yann-Hang (Committee member) / Bolotin, Evgeny (Committee member) / Arizona State University (Publisher)
Created2018
Description
This dissertation describes a process for interface capturing via an arbitrary-order, nearly quadrature free, discontinuous Galerkin (DG) scheme for the conservative level set method (Olsson et al., 2005, 2008). The DG numerical method is utilized to solve both advection and reinitialization, and executed on a refined level set grid (Herrmann, 2008) for effective use of processing power. Computation is executed in parallel utilizing both CPU and GPU architectures to make the method feasible at high order. Finally, a sparse data structure is implemented to take full advantage of parallelism on the GPU, where performance relies on well-managed memory operations.
With solution variables projected into a kth order polynomial basis, a k+1 order convergence rate is found for both advection and reinitialization tests using the method of manufactured solutions. Other standard test cases, such as Zalesak's disk and deformation of columns and spheres in periodic vortices are also performed, showing several orders of magnitude improvement over traditional WENO level set methods. These tests also show the impact of reinitialization, which often increases shape and volume errors as a result of level set scalar trapping by normal vectors calculated from the local level set field.
Accelerating advection via GPU hardware is found to provide a 30x speedup factor comparing a 2.0GHz Intel Xeon E5-2620 CPU in serial vs. a Nvidia Tesla K20 GPU, with speedup factors increasing with polynomial degree until shared memory is filled. A similar algorithm is implemented for reinitialization, which relies on heavier use of shared and global memory and as a result fills them more quickly and produces smaller speedups of 18x.
With solution variables projected into a kth order polynomial basis, a k+1 order convergence rate is found for both advection and reinitialization tests using the method of manufactured solutions. Other standard test cases, such as Zalesak's disk and deformation of columns and spheres in periodic vortices are also performed, showing several orders of magnitude improvement over traditional WENO level set methods. These tests also show the impact of reinitialization, which often increases shape and volume errors as a result of level set scalar trapping by normal vectors calculated from the local level set field.
Accelerating advection via GPU hardware is found to provide a 30x speedup factor comparing a 2.0GHz Intel Xeon E5-2620 CPU in serial vs. a Nvidia Tesla K20 GPU, with speedup factors increasing with polynomial degree until shared memory is filled. A similar algorithm is implemented for reinitialization, which relies on heavier use of shared and global memory and as a result fills them more quickly and produces smaller speedups of 18x.
ContributorsJibben, Zechariah J (Author) / Herrmann, Marcus (Thesis advisor) / Squires, Kyle (Committee member) / Adrian, Ronald (Committee member) / Chen, Kangping (Committee member) / Treacy, Michael (Committee member) / Arizona State University (Publisher)
Created2015