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 13 of 13 for “"Network architecture design"”.
-
Automated autoencoder neural network architecture design
… large datasets. Autoencoders, a class of neural networks, have demonstrated significant potential in applications, such as dimensionality reduction, feature extraction, anomaly detection, thereby making them powerful tools for managing vast amounts of data. However, the manual design and …
-
Computational mathematics approaches to deep neural network architecture design
Designing appropriate neural network architectures remains one of the central open challenges in deep learning. The performance of a neural network depends critically on its depth, width, and connectivity. Yet identifying a suitable architecture for a given task is still largely ad hoc, relying on …
-
NEURAL NETWORK ARCHITECTURE DESIGN FOR MILITARY DECISION-MAKING UNDER FOG OF WAR
… Q-value functions implemented as deep neural networks to select the most promising action from a given state. Our approach trains Q-value approximators on perfect-information data to learn tactical decisions from observable features, then evaluates them under both perfect and imperfect …
-
Evolutionary multi-objective bi-level optimization for efficient deep neural network architecture design
Deep convolutional neural networks (CNNs) are the backbones of deep learning (DL) paradigms for numerous vision tasks, including object recognition, detection, segmentation, etc. Early advancements in CNN architectures are primarily driven by human expertise and elaborate design. Recently, neural …
-
EFFICIENT ARCHITECTURE DESIGN FOR DEEP NEURAL NETWORKS
Neural network architecture design is always one of the most attractive and important research topics in the era of deep learning. Early neural network design [112, 49, 71] mostly relies on standard convolutions and pooling techniques. Despite of achieving good performance, their learnable …
-
Learning invariant representations in neural networks
… learning tasks. While convolutional neural networks famously and successfully exploit translational symmetries, other symmetries have, until recently, often been neglected. Incorporating symmetries or invariances into the neural network architecture avoids costly data augmentation and …
-
Cancer Detection System Using Binary Neural Network on DNA Based Architecture
… can be further scaled to support Binary Neural Network (BNN) models, which are particularly well-suited for mitigating stochastic noise at the molecular level. In recent years, clinical studies have established that as tumor cell populations expand, the shedding of circulating tumor DNA (ctDNA) …
-
Advancing Deep Learning for Data and Signal Processing Applications
… of realistic, standardized benchmark datasets, architectures that do not fully exploit the structure of domain-specific data, slow or resource-inefficient training procedures, and a limited theoretical understanding of the fundamental limits of learning algorithms. This dissertation addresses …
-
3D Pose and shape estimation of hands and manipulated objects from images and videos
… strategy in which two branches of deep neural network mutually exchange information for 3D hand-object reconstruction from single RGB image. Second, we present a new Transformer-based method that estimates the absolute root pose and shape of two-hands with extended forearm at high resolution …
-
Building efficient wireless infrastructures for pervasive computing environments
… these devices usually are not isolated, but networked with each other through wireless channels so that people can easily control and access them. In the architecture of pervasive computing systems, these small and networked computing devices form a wireless infrastructure layer to support …
-
On neural networks with equivariance or invariance property
… perceptrons, commonly known as neural networks, are universal approximators that can approximate any continuous functions arbitrarily well when the number of perceptrons is allowed to grow indefinitely. When prior knowledge about the target function is available, constraints can be …
-
Εξέλιξη πρωτοκόλλου SCP-ECG για μεταφορά βιοσημάτων πολλαπλών τύπων σε ιατρικά πληροφοριακά συστήματα: υλοποίηση πιλοτικού τηλεϊατρικού συστήματος
… geolocation data, etc.), through a communication network to a Health Reception Center. The dissertation also defines the architecture of a Health Tele-monitoring System (HTS) aiming to protocol’s application and evaluation. The pilot HTS, uses the protocol e-SCP-ECG+, in order to transmit, manage …
-
Σχεδίαση και ανάπτυξη επικοινωνιακής αρχιτεκτονικής συνδυασμένων επιπέδων σε κατανεμημένα ασύρματα δίκτυα αισθητήρων με απαιτήσεις απόκρισης πραγματικού χρόνου
… on the development of distributed wireless networks with real-time response demands and scarce available resources. Furthermore, another main goal is the design and proposal of a respective network architecture aiming at optimum resource management under congestion scenarios and therefore …