Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 5 of 5
Full-Text Articles in Engineering
Bayesian Convolutional Neural Network With Prediction Smoothing And Adversarial Class Thresholds, Noah M. Miller
Bayesian Convolutional Neural Network With Prediction Smoothing And Adversarial Class Thresholds, Noah M. Miller
Theses and Dissertations
Using convolutional neural networks (CNNs) for image classification for each frame in a video is a very common technique. Unfortunately, CNNs are very brittle and have a tendency to be over confident in their predictions. This can lead to what we will refer to as “flickering,” which is when the predictions between frames jump back and forth between classes. In this paper, new methods are proposed to combat these shortcomings. This paper utilizes a Bayesian CNN which allows for a distribution of outputs on each data point instead of just a point estimate. These distributions are then smoothed over multiple …
The Autonomous Attack Aviation Problem, John C. Goodwill
The Autonomous Attack Aviation Problem, John C. Goodwill
Theses and Dissertations
An autonomous unmanned combat aerial vehicle (AUCAV) performing an air-to-ground attack mission must make sequential targeting and routing decisions under uncertainty. We formulate a Markov decision process model of this autonomous attack aviation problem (A3P) and solve it using an approximate dynamic programming (ADP) approach. We develop an approximate policy iteration algorithm that implements a least squares temporal difference learning mechanism to solve the A3P. Basis functions are developed and tested for application within the ADP algorithm. The ADP policy is compared to a benchmark policy, the DROP policy, which is determined by repeatedly solving a deterministic orienteering problem as …
Meta-Heuristic Optimization Methods For Quaternion-Valued Neural Networks, Jeremiah P. Bill
Meta-Heuristic Optimization Methods For Quaternion-Valued Neural Networks, Jeremiah P. Bill
Theses and Dissertations
In recent years, real-valued neural networks have demonstrated promising, and often striking, results across a broad range of domains. This has driven a surge of applications utilizing high-dimensional datasets. While many techniques exist to alleviate issues of high-dimensionality, they all induce a cost in terms of network size or computational runtime. This work examines the use of quaternions, a form of hypercomplex numbers, in neural networks. The constructed networks demonstrate the ability of quaternions to encode high-dimensional data in an efficient neural network structure, showing that hypercomplex neural networks reduce the number of total trainable parameters compared to their real-valued …
Predicting Upper Atmospheric Weather Conditions Utilizing Long-Short Term Memory Neural Networks For Aircraft Fuel Efficiency, Garrett A. Alarcon
Predicting Upper Atmospheric Weather Conditions Utilizing Long-Short Term Memory Neural Networks For Aircraft Fuel Efficiency, Garrett A. Alarcon
Theses and Dissertations
Aviation fuel is a major component of the Air Force (AF) budget, and vital for the core mission of the AF. This study investigated the viability of LSTMs to increase the accuracy of deterministic NWP models, while also investigating the ability to reduce model generation time. Increased forecast accuracy for wind speeds could be implemented into existing flight path models to further increase fuel efficiency, while reduced modeling times would allow flight planners to generate a flight plan in rapid response situations. The most viable model consisted of an ensemble of six LSTMs trained o six coordinates. The model's error …
Modeling Small Unmanned Aerial System Mishaps Using Logistics Regression And Artificial Neural Networks, Sean E. Wolf
Modeling Small Unmanned Aerial System Mishaps Using Logistics Regression And Artificial Neural Networks, Sean E. Wolf
Theses and Dissertations
A dataset of 854 small unmanned aerial system (SUAS) flight experiments from 2005-2009 is analyzed to determine significant factors that contribute to mishaps. The data from 29 airframes of different designs and technology readiness levels were aggregated. 20 measured parameters from each flight experiment are investigated, including wind speed, pilot experience, number of prior flights, pilot currency, etc. Outcomes of failures (loss of flight data) and damage (injury to airframe) are classified by logistic regression modeling and artificial neural network analysis. From the analysis, it can be concluded that SUAS damage is a random event that cannot be predicted with …