Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 36 for “"Neural Network Optimization"”.

  1. Adaptive Momentum for Neural Network Optimization

    … a novel and efficient algorithm for optimizing neural networks inspired by a recently proposed geodesic optimization algorithm. Our algorithm, which we call Stochastic Geodesic Optimization (SGeO), utilizes an adaptive coefficient on top of Polyaks Heavy Ball method effectively controlling the …

    york Repository record for Adaptive Momentum for Neural Network Optimization (opens in a new tab)

  2. A Hamiltonian Systems Approach to Neural Network Optimization

    … structure-preserving methods for first-order optimization of Lipschitz smooth objectives by interpreting the dynamics as a dissipative Hamiltonian system, in which the model parameters evolve jointly with an auxiliary momentum variable. This formulation induces a natural energy dissipation …

    waterloo Repository record for A Hamiltonian Systems Approach to Neural Network Optimization (opens in a new tab)

  3. Theoretical Understanding of Neural Network Optimization Landscape and Self-Supervised Representation Learning

    <p>Neural networks have achieved remarkable empirical success in various areas. One key factor of their success is their ability to automatically learn useful representations from data. Self-supervised representation learning, which learns the representations during pre-training and applies learned …

    duke Repository record for Theoretical Understanding of Neural Network Optimization Landscape and Self-Supervised Representation Learning (opens in a new tab)

  4. Convolutional Neural Network Optimization and Parallel Compressive Sensing Algorithms for Accelerated MRI Reconstruction

    … learning-based MRI reconstruction methods, the network’s architecture plays a crucial role in the quality of the reconstructed image. According to the large number of architectures that can be considered for these models, manually designing architectures and testing all the possible solutions …

    tamu Repository record for Convolutional Neural Network Optimization and Parallel Compressive Sensing Algorithms for Accelerated MRI Reconstruction (opens in a new tab)

  5. DEVELOPMENT OF A 3D SIMULATION TOOL LEVERAGING GENETIC ALGORITHMS FOR NEURAL NETWORK OPTIMIZATION IN AUTOMATED DRIVING SYSTEMS: AN APPLICATION FOR SCENARIO ANALYSIS

    … (GA) with selection methods and artificial neural network (ANN) architectures all configurable via a user interface (UI), which guides the training and optimization of ANNs, allowing ADS policies to learn and evolve rather than relying on existing datasets. The Unity game engine provides the …

    jhu Repository record for DEVELOPMENT OF A 3D SIMULATION TOOL LEVERAGING GENETIC ALGORITHMS FOR NEURAL NETWORK OPTIMIZATION IN AUTOMATED DRIVING SYSTEMS: AN APPLICATION FOR SCENARIO ANALYSIS (opens in a new tab)

  6. Deep Learning Optimization and Acceleration

    The novelty of this dissertation is the optimization and acceleration of deep neural networks aimed at real-time predictions with minimal energy consumption. It consists of cross-layer optimization, output directed dynamic quantization, and opportunistic near-data computation for deep neural …

    unt Repository record for Deep Learning Optimization and Acceleration (opens in a new tab)

  7. Completion of hinge loss has an implicit bias

    … gradient descent iterates for two-layer linear networks on the inner product loss function is derived. Using the derived dynamical system, a precise explicit algorithm is developed which when implemented reproduces the gradient descent iterates of two-layer ReLU nets on the inner product …

    uiuc Repository record for Completion of hinge loss has an implicit bias (opens in a new tab)

  8. Optimizing neural networks through activation function discovery and automatic weight initialization

    … present methods focus on hyperparameters and neural network topologies, other aspects of neural network design can be optimized as well. To further the state of the art in AutoML, this dissertation introduces techniques for discovering more powerful activation functions and establishing more …

    texas Repository record for Optimizing neural networks through activation function discovery and automatic weight initialization (opens in a new tab)

  9. QUANTIZATION AND APPROXIMATION OF NEURAL NETWORKS

    … specific focus on the theoretical properties of neural networks. The first part of this work addresses neural network quantization. Despite the widespread use of quantization as a primary method for compressing deep neural networks, a majority of existing approaches lack provable guarantees. By …

    maryland Repository record for QUANTIZATION AND APPROXIMATION OF NEURAL NETWORKS (opens in a new tab)

  10. Text-Independent Automatic Speaker Identification Using Partitioned Neural Networks

    … which is applied to talker identification using neural networks. In recent years artificial neural networks have been shown to work exceptionally well for small but difficult pattern classification tasks. However, their application to large tasks (i.e., having more than ten to 20 categories) is …

    odu Repository record for Text-Independent Automatic Speaker Identification Using Partitioned Neural Networks (opens in a new tab)

  11. Study of Implementation of CNN on Low-power Platform for Smart Traffic Optimization

    … the traffic in a smarter way. Since the deep neural network has shown its great value in vehicle detection area, people may wonder what kind of impact would be brought by the combination of IoT and deep learning techniques. In this work, an exploration of implementation of CNN (convolutional …

    waterloo Repository record for Study of Implementation of CNN on Low-power Platform for Smart Traffic Optimization (opens in a new tab)

  12. Robust actor-critic trajectory optimization and Autonomous harvesting robot integration

    … is organized into two parts. First, we develop a neural network–based constrained trajectory optimization algorithm with stability guarantees, along with a neural network–driven numerical error correction. The framework reformulates the structure of an optimal control problem as a neural network …

    ucf Repository record for Robust actor-critic trajectory optimization and Autonomous harvesting robot integration (opens in a new tab)

  13. Control Power Optimization using Artificial Intelligence for Hybrid Wing Body Aircraft

    Traditional methods of control allocation optimization have shown difficulties in exploiting the full potential of controlling a large array of control surfaces. This research investigates the potential of employing artificial intelligence methods like neurocomputing to the control allocation …

    vt Repository record for Control Power Optimization using Artificial Intelligence for Hybrid Wing Body Aircraft (opens in a new tab)

  14. Mean Field Langevin Dynamics, Mean Field Neural Networks, and Mean Field Ising Models

    … understanding of Markov chain mixing, mean field optimization, and neural network training. First, we introduce a virtual particle stochastic approximation algorithm for mean field Langevin dynamics. The key innovation is a two-particle system: real particles that form the output and virtual …

    oregon Repository record for Mean Field Langevin Dynamics, Mean Field Neural Networks, and Mean Field Ising Models (opens in a new tab)

  15. Causally-Guided Evolutionary Computation for Design

    … evolutionary processes is demonstrated on a neural network optimization task. These methods are then extended to include crossover, and the synergistic effects of causally-guided crossover and mutation are demonstrated when applied to a real-world antenna design task. Causally-guided mutation …

    maryland Repository record for Causally-Guided Evolutionary Computation for Design (opens in a new tab)

  16. Probabilistic Calibration of Unreinforced Masonry Wall Properties: From Constitutive Material Models to Structural Performance

    … material parameters using random variables. A Neural Network Optimization algorithm is proposed to identify and characterize the material parameters for describing the mechanical response of unreinforced masonry walls. Unreinforced masonry wall models are validated using the optimal material …

    tamu Repository record for Probabilistic Calibration of Unreinforced Masonry Wall Properties: From Constitutive Material Models to Structural Performance (opens in a new tab)

  17. Architecture for intelligent power systems management, optimization, and storage.

    The management of power and the optimization of systems generating and using power are critical technologies. A new architecture is developed to advance the current state of the art by providing an intelligent and autonomous solution for power systems management. The architecture is two-layered and …

    louisville Repository record for Architecture for intelligent power systems management, optimization, and storage. (opens in a new tab)

  18. DEEP UNSUPERVISED MODELS LEVERAGING LEARNING AND REASONING

    … a general unsupervised framework (deep reasoning networks, DRNets) for demixing and inferring crystal structure. DRNets seamlessly integrate reasoning about prior scientific knowledge into neural network optimization using an interpretable latent space. As a result, DRNets require only modest …

    cornell Repository record for DEEP UNSUPERVISED MODELS LEVERAGING LEARNING AND REASONING (opens in a new tab)

  19. Combining Deep Learning with Reasoning: from Mapping Species to Solving Games and Crystal Structures

    … constraint reasoning seamlessly integrated into neural network optimization. We first illustrate this idea in the context of supervised learning on multi-label classification --- in particular, on joint species distribution modeling, where we propose the Deep Multivariate Probit Model (DMVP) to …

    cornell Repository record for Combining Deep Learning with Reasoning: from Mapping Species to Solving Games and Crystal Structures (opens in a new tab)

  20. An Edge Computing Device Optimized and Transfer Learning Enhanced Deep Learning Model for Detecting Wildfire Flame and Smoke

    … to two deployment optimized frameworks, Open Neural Network Exchange (ONNX) and Open Visual Inference and Neural Network Optimization (OpenVINO), to accelerate CPU-based inference. Compared to the original YOLOv5n, the modified model converted to OpenVINO demonstrates a 423% increase in …

    unlv Repository record for An Edge Computing Device Optimized and Transfer Learning Enhanced Deep Learning Model for Detecting Wildfire Flame and Smoke (opens in a new tab) Download PDF for An Edge Computing Device Optimized and Transfer Learning Enhanced Deep Learning Model for Detecting Wildfire Flame and Smoke (opens in a new tab)

Page 1 of 2