Theses - Computation for Design and Optimization
http://hdl.handle.net/1721.1/39116
2016-07-12T06:09:30ZAutomatic design tool for robust radio frequency decoupling matrices in magnetic resonance imaging
http://hdl.handle.net/1721.1/97791
Automatic design tool for robust radio frequency decoupling matrices in magnetic resonance imaging
Mahmood, Zohaib
In this thesis we study the design of robust decoupling matrices for coupled transmit radio frequency arrays used in magnetic resonance imaging (MRI). In a coupled parallel transmit array, because of the coupling itself, the power delivered to a channel is typically partially re-distributed to other channels. This power must then be dissipated in circulators resulting into a significant reduction in the power efficiency of the overall system. In this thesis, we propose an automated approach to design a robust decoupling matrix interfaced between the RF amplifiers and the coils. The decoupling matrix is optimized to ensure all forward power is delivered to the load. The decoupling condition dictates that the admittance matrix seen by power amplifiers with 50 Ohms output impedance is a diagonal matrix with matching 1 (or 0.02 Siemens) at the diagonal. Our tool computes the values of the decoupling matrix via a non linear optimization and generate a physical realization using reactive elements such as inductors and capacitors. The methods presented in this thesis scale to any arbitrary number of channels and can be readily applied to other coupled systems such as antenna arrays. Furthermore our tool computes parameterized dynamical models and performs sensitivity analysis with respect to patient head-size and head-position for MRI coils.
Thesis: S.M., Massachusetts Institute of Technology, Computation for Design and Optimization Program, 2015.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 43-45).
2015-01-01T00:00:00ZOptimal Bayesian experimental design in the presence of model error
http://hdl.handle.net/1721.1/97790
Optimal Bayesian experimental design in the presence of model error
Feng, Chi, S.M. Massachusetts Institute of Technology
The optimal selection of experimental conditions is essential to maximizing the value of data for inference and prediction. We propose an information theoretic framework and algorithms for robust optimal experimental design with simulation-based models, with the goal of maximizing information gain in targeted subsets of model parameters, particularly in situations where experiments are costly. Our framework employs a Bayesian statistical setting, which naturally incorporates heterogeneous sources of information. An objective function reflects expected information gain from proposed experimental designs. Monte Carlo sampling is used to evaluate the expected information gain, and stochastic approximation algorithms make optimization feasible for computationally intensive and high-dimensional problems. A key aspect of our framework is the introduction of model calibration discrepancy terms that are used to "relax" the model so that proposed optimal experiments are more robust to model error or inadequacy. We illustrate the approach via several model problems and misspecification scenarios. In particular, we show how optimal designs are modified by allowing for model error, and we evaluate the performance of various designs by simulating "real-world" data from models not considered explicitly in the optimization objective.
Thesis: S.M., Massachusetts Institute of Technology, Computation for Design and Optimization Program, 2015.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 87-90).
2015-01-01T00:00:00ZAn analysis on information diffusion by retweets in Twitter
http://hdl.handle.net/1721.1/97348
An analysis on information diffusion by retweets in Twitter
Sakamoto, Tomoaki
This dissertation examines retweeting activities as the information spreading function of Twitter. First, we investigated what kind of features of a tweet help to get retweets. We construct a model that describes peoples' decision making on retweets, and with related observation, we show that more retweeted tweets get retweeted more. In terms of specific features of tweets, it has been shown that the number of followers and the number of retweets are positively correlated, and hashtags attract more retweets than the tweets without hashtags. On the other hand, we also found that including hashtags and getting one or more retweets are statistically independent. Moreover, we showed including URLs or user-mentions in tweets and getting one or more retweets are statistically independent. In our results, including a picture is slightly effective to get this sense of retweetability. Second, we compare the retweeters of tweets including a picture and only text, especially focusing on distance from the original tweeters. Comparing the ratio of retweets by followers of the author of the original tweets among the initial 50 retweets, tweets with a picture have a slightly lower ratio, though there is no significant difference between the average for tweets with pictures and without pictures at the 95% significance level. We also investigate how many retweets are posted by users in followers' network connected to the original tweeter, and show that the depths of retweeters' network for tweets with picture have larger variance than that of tweets without pictures. This result implies that a tweet including picture can reach more people than a tweet without a picture potentially.
Thesis: S.M., Massachusetts Institute of Technology, Computation for Design and Optimization Program, 2015.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 81-82).
2015-01-01T00:00:00ZEnergy optimal path planning using stochastic dynamically orthogonal level set equations
http://hdl.handle.net/1721.1/95564
Energy optimal path planning using stochastic dynamically orthogonal level set equations
Narayanan Subramani, Deepak
The growing use of autonomous underwater vehicles and underwater gliders for a variety of applications gives rise to new requirements in the operation of these vehicles. One such important requirement is optimization of energy required for undertaking missions that will enable longer endurance and lower operational costs. Our goal in this thesis is to develop a computationally efficient, and rigorous methodology that can predict energy-optimal paths from among all time-optimal paths to complete an underwater mission. For this, we develop rigorous a new stochastic Dynamically Orthogonal Level Set optimization methodology. In our thesis, after a review of existing path planning methodologies with a focus on energy optimality, we present the background of time-optimal path planning using the level set method. We then lay out the questions that inspired the present thesis, provide the goal of the current work and explain an extension of the time-optimal path planning methodology to the time-optimal path planning in the case of variable nominal engine thrust. We then proceed to state the problem statement formally. Thereafter, we develop the new methodology for solving the optimization problem through stochastic optimization and derive new Dynamically Orthogonal Level Set Field equations. We then carefully present different approaches to handle the non-polynomial non-linearity in the stochastic Level Set Hamilton-Jacobi equations and also discuss the computational efficiency of the algorithm. We then illustrate the inner-workings and nuances of our new stochastic DO level set energy optimal path planning algorithm through two simple, yet important, canonical steady flows that simulate a stead front and a steady eddy. We formulate a double energy-time minimization to obtain a semi-analytical energy optimal path for the steady front crossing test case and compare the results to these of our stochastic DO level set scheme. We then apply our methodology to an idealized ocean simulation using Double Gyre flows, and finally show an application with real ocean data for completing a mission in the Middle Atlantic Bight and New Jersey Shelf/Hudson Canyon region.
Thesis: S.M., Massachusetts Institute of Technology, Computation for Design and Optimization Program, 2014.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 93-100).
2014-01-01T00:00:00Z