DSpace at MIT
This site is a university repository providing access to the publication output of the institution. Registered users can set up email alerts to notify them of newly added relevant content. A certain level of encryption and security is embedded in the site which may cause some users accessibility problems.
Electrical Engineering and Computer Sciences - Master's degree
Mostrando recursos 1 - 20 de 5,490
Cyber safety : a systems thinking and systems theory approach to managing cyber security risks - Salim, Hamid M
If we are to manage cyber security risks more effectively in today's complex and dynamic Web 2.0 environment, then a new way of thinking is needed to complement traditional approaches. According to Symantec's 2014 Internet Security Threat Report, in 2012 more than ten million identities that included real names, dates of birth, and social security were exposed by a single breach. In 2013 there were eight breaches that each exposed over ten million identities. These breaches were recorded despite the fact that significant resources are expended, on managing cyber security risks each year by businesses and governments. The objective of...
Stochastic capacity modeling to support demand/capacity gap planning - Niles, Augusta (Augusta L.)
Capacity strategy has established methods of dealing with uncertainty in future demand. This project advances the concept of capacity strategy under conditions of uncertainty in cases where capacity is the primary source of uncertainty. Novartis Vaccines, one of five divisions of Novartis AG, produces nearly two dozen vaccines which are offered in syringes, vials, multi or single pack, and multi or single dose and delivered in language-specific packaging to countries all over the world. Bexsero is a new product in 2013. As demand for Bexsero and other products increases over the next ten years, the production lines used to package...
Modeling neuroscience patient flow and inpatient bed management - Hiltrop, Jonas
Massachusetts General Hospital (MGH) experiences consistently high demand for its more than 900 inpatient beds. On an average weekday, the hospital admits about 220 patients, with the emergency department (ED) and the operating rooms (OR) being the main sources of admissions. Given MGH's high occupancy rates, a comparable number of discharges have to occur daily, and the intraday time distributions of admissions and discharges have to be aligned in order to avoid long wait times for beds. The situation is complicated by the specialization of beds and the medical needs of patients, which place constraints on the possible bed-patient assignments....
Profitability improvement through profit mapping - Chang, Brian, S.M. Massachusetts Institute of Technology
Faced with a dynamically changing market, Boston Scientific's Cardiac Rhythm Management is seeking ways to increase its profitability. The division is experiencing a phenomenon where pulse generators and leads' product life cycle transitions may be slower than ideal. In addition, Boston Scientific is exploring ways to systematically and accurately assess its products' return on investments. The approach to the project is a technique called Profit Mapping. By carefully assessing the true full costs of each transaction, we can determine the actual profitability of each account and product and use the visualized results to detect anomalies, test hypotheses, and formulate action...
Optimal allocation of surgical services - Braun, Marcus (Marcus D.)
Over the past several years Greater Boston has witnessed the consolidation of multiple community hospitals into larger care organizations and a renewed focus on the delivery of affordable care. In order for the Beth Israel Deaconess Medical Center (BIDMc) to respond and adapt to this changing landscape it will be critical to not only understand demand and capacity across the organization's entire network, but also to recognize how the deployment of limited resources can best be improved. From a BIDMc Department of Surgery Perspective, essential business questions include: 1 How to allocate limited existing resources efficiently? 2 Which future growth...
Physical modeling of wind turbine generators in a small scale analog system - Wang, Xuntuo
This project represents the physical modeling and experimental test of a Doubly-fed Induction Machine (DFIM), in order to substantially analyze the characteristic behaviors of wind turbines and its use in the micro-grid network. The environment set-up is based on a smart micro-grid system, which consists of the grid, a diesel generator, a solar panel, and a wind farm. The hardware work includes the design of a 2.5kW inverter and a L-C-L grid filter (including inductor design and construction). The goal of this research is to better emulate the operation principles of wind turbine system. Future work proposes developing a better...
Visualizing database queries - Vartak, Manasi
Data analysts operating on large volumes of data often rely on visualizations to interpret the results of queries. However, finding the right visualization for a query is a laborious and time-consuming task. We propose SEEDB, a system that partially automates this task: given a query, SEEDB explores the space of all possible visualizations, and automatically identifies and recommends to the analyst those visualizations it finds to be most "interesting" or "useful".
Computational 3D and reflectivity imaging with high photon efficiency - Shin, Dongeek
Imaging the 3D structure and reflectivity of a scene can be done using photon-counting detectors. Traditional imagers of this type typically require hundreds of detected photons per pixel for accurate 3D and reflectivity imaging. Under low light-level conditions, in which the mean photon count is small, the inverse problem of forming 3D and reflectivity images is difficult due to the Poisson noise inherent in low-flux operation. In this thesis, we propose and study two computational imagers (one passive, one active) that can form accurate images at low light levels. We demonstrate the superior imaging quality of the proposed imagers by...
Continuous and non-invasive blood pressure monitoring using ultrasonic methods - Seo, Joohyun
This thesis presents a continuous and non-invasive arterial blood pressure (CNAP) monitoring technique using ultrasound. An arterial blood pressure (ABP) waveform provides valuable information in treating cardiovascular diseases. Although an invasive ABP measurement through arterial catheterization performed in an intensive care unit (ICU) is considered a gold standard, its invasive nature not only increases various patients' risks but makes its usage for cardiovascular studies expensive. Therefore, reliable non-invasive ABP waveform estimation has been desired for a long time by medical communities. This work details ABP waveform estimation based on a vessel cross-sectional area measurement combined with the elastic property of...
A comparative analysis of physical-layer rateless coding architectures - Romero, David Luis
An analysis of rateless codes implemented at the physical layer is developed. Our model takes into account two aspects of practical communication system design that are abstracted away in many existing works on the subject. In particular, our model assumes that : (1) practical error detection methods are used to determine when to terminate decoding; and (2) performance and reliability as observed at the transport layer are the metrics of interest. Within the context of these assumptions, we then evaluate two recently proposed high-performing rateless codes. Using our analysis to guide an empirical study, the process of selecting the best...
Morphological segmentation : an unsupervised method and application to Keyword Spotting - Narasimhan, Karthik Rajagopal
The contributions of this thesis are twofold. First, we present a new unsupervised algorithm for morphological segmentation that utilizes pseudo-semantic information, in addition to orthographic cues. We make use of the semantic signals from continuous word vectors, trained on huge corpora of raw text data. We formulate a log-linear model that is simple and can be used to perform fast, efficient inference on new words. We evaluate our model on a standard morphological segmentation dataset, and obtain large performance gains of up to 18.4% over an existing state-of-the-art system, Morfessor. Second, we explore the impact of morphological segmentation on the...
Low threshold organic and quantum dot nanobeam lasers - Mahony, Thomas Stephen
Lasers in the visible spectrum have many applications including sensing, medical, and entertainment applications. Traditional semiconductors face challenges that limit their ability to create lasers for the visible spectrum. Organic materials and quantum (lots are an attractive alternative for visible lasers due to their broad, tunable emission and deposition using fabrication techniques of low complexity. These materials have been used to demonstrate low lasing thresholds, and we hope to improve upon them with a novel design, paving the way towards electrically pumped and continuous wave operation lasing. In this thesis we couple the use of organic materials and quantum dots...
Nano-scale ohmic contacts for III-V MOSFETs - Lu, Wenjie
As modem silicon CMOS has been scaled down to extremely small dimensions, there is an urgent need for technological innovations of new devices architectures that would allow the continuation of Moore's Law into the future. In particular, for CMOS with nanometer scale pitch size, the intrinsic electronic properties of silicon as channel material represent a significant hindrance to further scaling. As a result, new channel materials are being investigated all over the world that would enable the push into the sub-10 nm regime. Among them, certain III-V compound semiconductors have emerged as the most promising candidates to replace silicon in...
Interpreting human activity from electrical consumption data through non-intrusive load monitoring - Gillman, Mark Daniel
Non-intrusive load monitoring (NILM) has three distinct advantages over today's smart meters. First, it offers accountability. Few people know where their kWh's are going. Second, it is a maintenance tool. Signs of wear are detectable through their electrical signal. Third, it provides awareness of human activity within a network. Each device has an electrical fingerprint, and specific devices imply associated human actions. From voltage and current measurements at a single point on the network, non-intrusive load monitoring (NILM) disaggregates appliance-level information. This information is available remotely in bandwidth-constrained environments. Four real-world field tests at military micro grids and commercial buildings...
Design of a novel test bench for induction heating load characterization - Fernandez del Castillo, Lisa
Magnetic materials used in induction heating applications have nonlinear magnetic properties with respect to field strength and frequency, which can be effectively characterized using experimental techniques. To this end, we present a test bench inverter optimized for induction heating experimentation, capable of driving an inductive load across a 1-100 kHz frequency range with up to 2 kW power. Harmonic distortion of the inverter is minimized with a novel multilevel topology and modulation scheme, thus allowing near-sinusoidal excitations to be obtained at varying field strengths and frequencies. To demonstration the capabilities of the test bench, we characterize the power dissipation of...
A programmable pipeline for multi-material fabrication - Vidimče, Kiril
3D printing hardware is rapidly scaling up to output continuous mixtures of multiple materials at increasing resolution over ever larger print volumes. This poses an enormous computational challenge: large high-resolution prints comprise trillions of voxels and petabytes of data and simply modeling and describing the input with spatially-varying material mixtures at this scale is challenging. Existing 3D printing software is insufficient; in particular, most software is designed to support only a few million primitives, with discrete material choices per object. In this body of work I present OpenFab, a programmable pipeline for synthesizing multi-material 3D printed objects that is inspired...
Beyond locality-sensitive hashing - Razenshteyn, Ilya
We present a new data structure for the c-approximate near neighbor problem (ANN) in the Euclidean space ... This is the first improvement over the result by Andoni and Indyk (FOCS 2006) and the first data structure that bypasses a locality-sensitive hashing lower bound proved by O'Donnell, Wu and Zhou (ICS 2011). By known reductions we obtain a data structure for the Hamming space and ... which is the first improvement over the result of Indyk and Motwani (STOC 1998).
Learning articulated motions from visual demonstration - Pillai, Sudeep
Robots operating autonomously in household environments must be capable of interacting with articulated objects on a daily basis. They should be able to infer each object's underlying kinematic linkages purely by observing its motion during manipulation. This work proposes a framework that enables robots to learn the articulation in objects from user-provided demonstrations, using RGB-D sensors. We introduce algorithms that combine concepts in sparse feature tracking, motion segmentation, object pose estimation, and articulation learning, to develop our proposed framework. Additionally, our methods can predict the motion of previously seen articulated objects in future encounters. We present experiments that demonstrate the...
Rare-earth doped aluminum oxide lasers for silicon photonics - Magden, Emir Salih
A reliable and CMOS-compatible deposition process for amorphous Al2O3 based active photonic components has been developed. Al2O3 films were reactively sputtered, where process optimization was achieved at a temperature of 250°C, with a deposition rate of 8.5 nm/min. With a surface roughness of 0.3 nm over a 1 [mu]m2 area, background optical losses as low as 0.1 dB/cm were obtained for undoped films. The development of active photonics components has been realized by use of rare-earth metals as dopants. By co-sputtering aluminum and erbium targets, Er3+ dopants at concentrations on the order of 1.0x1020 cm-3 have been added to the...