Current Search:  Simulated annealing Mathematics (x)

View All Items

  • CSV Spreadsheet
(1 - 3 of 3)
APPLYING BLIND SOURCE SEPARATION TO MAGNETIC ANOMALY DETECTION
Title
APPLYING BLIND SOURCE SEPARATION TO MAGNETIC ANOMALY DETECTION.
Creator
Nieves, Eric, Beaujean, Pierre-Philippe, Florida Atlantic University, Department of Ocean and Mechanical Engineering, College of Engineering and Computer Science
Abstract/Description
The research shows a novel approach for the Magnetic Anomaly Differentiation and Localization Algorithm, which simultaneously localizes multiple magnetic anomalies with weak total field signatures (tens of nT). In particular, it focuses on the case where there are two homogeneous targets with known magnetic moments. This was done by analyzing the magnetic signals and adapting Independent Component Analysis (ICA) and Simulated Annealing (SA) to solve the problem statement. The results show the...
Show moreThe research shows a novel approach for the Magnetic Anomaly Differentiation and Localization Algorithm, which simultaneously localizes multiple magnetic anomalies with weak total field signatures (tens of nT). In particular, it focuses on the case where there are two homogeneous targets with known magnetic moments. This was done by analyzing the magnetic signals and adapting Independent Component Analysis (ICA) and Simulated Annealing (SA) to solve the problem statement. The results show the groundwork for using a combination of fastICA and SA to give localization errors of 3 meters or less per target in simulation and achieved a 58% success rate. Experimental results experienced additional errors due to the effects of magnetic background, unknown magnetic moments, and navigation error. While one target was localized within 3 meters, only the latest experimental run showed the second target approaching the localization specification. This highlighted the need for higher signal-to-noise ratio and equipment with better navigational accuracy. The data analysis was used to provide recommendations on the needed equipment to minimize observed errors and improve algorithm success.
Show less
Date Issued
2020
PURL
http://purl.flvc.org/fau/fd/FA00013610
Subject Headings
Magnetic anomalies, Simulated annealing (Mathematics), Independent component analysis, Unmanned vehicles
Format
Document (PDF)
A GPU- BASED SIMULATED ANNEALING ALGORITHM FOR INTENSITY-MODULATED RADIATION THERAPY
Title
A GPU- BASED SIMULATED ANNEALING ALGORITHM FOR INTENSITY-MODULATED RADIATION THERAPY.
Creator
Galanakou, Panagiota, Leventouri, Theodora, Florida Atlantic University, Department of Physics, Charles E. Schmidt College of Science
Abstract/Description
Simulating Annealing Algorithm (SAA) has been proposed for optimization of the Intensity-Modulated Radiation Therapy (IMRT). Despite the advantage of the SAA to be a global optimizer, the SAA optimization of IMRT is an extensive computational task due to the large scale of the optimization variables, and therefore it requires significant computational resources. In this research we introduce a parallel graphics processing unit (GPU)-based SAA developed in MATLAB platform and compliant with...
Show moreSimulating Annealing Algorithm (SAA) has been proposed for optimization of the Intensity-Modulated Radiation Therapy (IMRT). Despite the advantage of the SAA to be a global optimizer, the SAA optimization of IMRT is an extensive computational task due to the large scale of the optimization variables, and therefore it requires significant computational resources. In this research we introduce a parallel graphics processing unit (GPU)-based SAA developed in MATLAB platform and compliant with the computational environment for radiotherapy research (CERR) for IMRT treatment planning in order elucidate the performance improvement of the SAA in IMRT optimization. First, we identify the “bottlenecks” of our code, and then we parallelize those on the GPU accordingly. Performance tests were conducted on four different GPU cards in comparison to a serial version of the algorithm executed on a CPU. A gradual increase of the speedup factor as a function of the number of beamlets was found for all four GPUs. A maximum speedup factor of 33.48 was achieved for a prostate case, and 30.51 for a lung cancer case when the K40m card and the maximum number of beams was utilized for each case. At the same time, the two optimized IMRT plans that were created (prostate and lung cancer plans) were met the IMRT optimization goals.
Show less
Date Issued
2019
PURL
http://purl.flvc.org/fau/fd/FA00013372
Subject Headings
Radiotherapy, Intensity-Modulated, Annealing algorithm, Simulated annealing (Mathematics), Graphics processing units
Format
Document (PDF)
FINANCIAL TIME-SERIES ANALYSIS WITH DEEP NEURAL NETWORKS
Title
FINANCIAL TIME-SERIES ANALYSIS WITH DEEP NEURAL NETWORKS.
Creator
Rimal, Binod, Hahn, William Edward, Florida Atlantic University, Department of Mathematical Sciences, Charles E. Schmidt College of Science
Abstract/Description
Financial time-series data are noisy, volatile, and nonlinear. The classic statistical linear models may not capture those underlying structures of the data. The rapid advancement in artificial intelligence and machine learning techniques, availability of large-scale data, and increased computational capabilities of a machine opens the door to developing sophisticated deep learning models to capture the nonlinearity and hidden information in the data. Creating a robust model by unlocking the...
Show moreFinancial time-series data are noisy, volatile, and nonlinear. The classic statistical linear models may not capture those underlying structures of the data. The rapid advancement in artificial intelligence and machine learning techniques, availability of large-scale data, and increased computational capabilities of a machine opens the door to developing sophisticated deep learning models to capture the nonlinearity and hidden information in the data. Creating a robust model by unlocking the power of a deep neural network and using real-time data is essential in this tech era. This study constructs a new computational framework to uncover the information in the financial time-series data and better inform the related parties. It carries out the comparative analysis of the performance of the deep learning models on stock price prediction with a well-balanced set of factors from fundamental data, macroeconomic data, and technical indicators responsible for stock price movement. We further build a novel computational framework through a merger of recurrent neural networks and random compression for the time-series analysis. The performance of the model is tested on a benchmark anomaly time-series dataset. This new computational framework in a compressed paradigm leads to improved computational efficiency and data privacy. Finally, this study develops a custom trading simulator and an agent-based hybrid model by combining gradient and gradient-free optimization methods. In particular, we explore the use of simulated annealing with stochastic gradient descent. The model trains a population of agents to predict appropriate trading behaviors such as buy, hold, or sell by optimizing the portfolio returns. Experimental results on S&P 500 index show that the proposed model outperforms the baseline models.
Show less
Date Issued
2022
PURL
http://purl.flvc.org/fau/fd/FA00014009
Subject Headings
Neural networks (Computer science), Deep learning (Machine learning), Time-series analysis, Stocks, Simulated annealing (Mathematics)
Format
Document (PDF)