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 4.715
Ecology and evolution simulation and quest design for an educational massive multiplayer online game - Zhang, Mark (Mark A.)
In this design-based research project, I developed two simulations to be used as student tools in a massively multiplayer online game targeted at STEM education, the Radix Endeavor. I designed both the underlying agent-based model as well as the user interface for each simulation, and furthermore designed quests for my simulations for the purposes of playtesting. My final ecological prototype is able to authentically model fairly complex food webs of six or more organisms, and my final evolutionary prototype can handle complex fitness relationships between the individual traits of a single population and various environmental factors. In my thesis, I...
Assessing intrusiveness of smartphone apps - Zhang, Fan, M. Eng. Massachusetts Institute of Technology
We tackle the challenge of improving transparency for smartphone apps by focusing on the intrusiveness component of assessing privacy risk. Specifically, we develop a framework for qualitatively assessing and quantitatively measuring the intrusiveness of apps based on their data access behavior. This framework has two essential components: 1) the Privacy Fingerprint, a concise yet holistic visual that captures the data access patterns unique to each app, including which types and under which privacy-relevant usage contexts sensitive data are collected, and 2) an Intrusiveness Score that numerically measures each app's level of intrusiveness, based on real data accesses gathered from empirical...
Evaluation of somewhat homomorphic encryption schemes - Yang, Yang, M. Eng. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Homomorphic encryption allows one to outsource expensive computation to an untrusted party while keeping data in an encrypted form. While there has been a growing research focus in fully homomorphic encryption schemes, many real-world applications require a scheme to be only "somewhat" homomorphic. Somewhat homomorphic encryption (SWHE) schemes, which support a limited number of homomorphic operations before encountering a decryption failure, are much more efficient than their fully homomorphic counterparts. In this thesis, I present the empirical evaluation of two SWHE scheme prototypes built for IARPA's Security and Privacy Assurance Research (SPAR) program. The evaluation captures the exact performance costs...
Reduced traces and JITing in Church - Wu, Jeff (Jeffrey K.)
Church is a Turing-complete probabilistic programming language, designed for inference. By allowing for easy description and manipulation of distributions, it allows one to describe classical Al models in compact ways, providing a language for very rich expression. However, for inference in Bayes nets, Hidden Markov Models, and topic models, the very settings for which probabilistic programming languages like Church were designed for, researchers typically instead write special cased algorithms in regular programming languages to maximize performance. In this paper, we argue that an extremely general language can still support very fast inference. We first introduce the theoretical aspects of our...
GPSZip : semantic representation and compression system for GPS using coresets - Wu, Cathy, M. Eng. Massachusetts Institute of Technology
We present a semantic approach for compressing mobile sensor data and focus on GPS streams. Unlike popular text-compression methods, our approach takes advantage of the fact that agents (robotic, personal, or vehicular) perform tasks in a physical space, and the resulting sensor stream usually contains repeated observations of the same locations, actions, or scenes. We model this sensor stream as a Markov process with unobserved states, and our goal is to compute the Hidden Markov Model (HMM) that maximizes the likelihood estimation (MLE) of generating the stream. Our semantic representation and compression system comprises of two main parts: 1) trajectory...
A controllable laser projector for diverting traffic - Wu, Brian Xiuken
For this Master's thesis, I designed and implemented a combined laser and controller system that can receive and convert commands from a computer into useful laser projections. In the last 10 years, on average, one police officer, who is conducting a vehicle stop or directing traffic, is struck and killed by an oncoming vehicle per month in the United States. Most of these instances occur at night. Current literature proposes that for some drivers, especially those who are fatigued, the combination of bright, rapidly flashing emergency lighting causes them to drive closer to a emergency vehicle rather than away from...
A material segmentation and classification system - Wong, Jennifer L
In this thesis, I developed a material segmentation and classification system that takes in images of an object and identifies the material composition of the object's surface. The 3D surface is first segmented into regions that likely contain the same material, using color as a heuristic measure. The material classification of each region is then based on the cosine lobe model. The cosine lobe model is our adopted reflectance model, which allows for a simple approximation of a material's reflectance properties, which then serves as the material's unique signature.
Optimizations in stream programming for multimedia applications - Wong, Eric, M. Eng. Massachusetts Institute of Technology
Multimedia applications are the most dominant workload in desktop and mobile computing. Such applications regularly process continuous sequences of data and can be naturally represented under the stream programming domain to take take advantage of domain-specific optimizations. Exploiting characteristics specific to multimedia programs can provide further significant impact on performance for this class of programs. This thesis identifies many multimedia applications that maintain induction variable state, which directly inhibits data parallelism for the program. We demonstrates it is essential to recognize and parallelize filters with induction variable state to enable scalable parallelization. We eliminate such state by introducing a new...
Competitive algorithms for online matching and vertex cover problems - Wong, Chiu Wai, M. Eng. Massachusetts Institute of Technology
The past decade has witnessed an explosion of research on the online bipartite matching problem. Surprisingly, its dual problem, online bipartite vertex cover, has never been explicitly studied before. One of the motivation for studying this problem is that it significantly generalizes the classical ski rental problem. An instance of such problems specifies a bipartite graph G = (L, R, E) whose left vertices L are offline and right vertices arrive online one at a time. An algorithm must maintain a valid vertex cover from which no vertex can ever be removed. The objective is to minimize the size of...
A ballistic transport model for HEMTs and III-V MOSFETs - Warnock, Shireen M
As silicon MOSFETs keep scaling down in size, the continued improvement on their logic performance is threated by their fundamental physical limits. With silicon approaching these limits, MOSFETs designed with III-V semiconductors have emerged as promising candidates to replace them. The low-effective mass of various III-V materials such as InGaAs and InAs allow both faster and more power efficient performance. One of the key challenges, particularly as devices continue to shrink, is to understand the important of non-idealities in FET structures. High-electron mobility transistors, or HEMTs, are III-V Quantum-Well FETs that we can use to explore many issues of relevance...
CharmMe : applying machine learning to facilitate meaningful interactions at the MIT Media Lab - Wang, Victor J
CharmMe is a social discovery application to help people connect with others of similar interests at a company, organization, or conference. Unlike traditional social networking or matching algorithms, CharmMe discovers connections automatically without the need for new profiles or tagging. By using natural language processing, we create a model of an organization by "reading" existing information related to the people being matched, such as their publications or social media accounts. Additionally, the application takes data provided by users Checking-in to conference talks or Liking projects, which are actions made popular by the social networking sites Facebook and Foursquare. To facilitate...
Minimizing interference due to genetic manipulation - Wang, Jerry S
Synthetic gene networks are the functional genetic 'programs' that will form the basis of increasingly sophisticated engineered organisms, cells and tissues. Important factors to consider in biologically engineering circuits are modularity and orthogonality of the components. We explore these factors experimentally by using TALEs as a synthetic transcription factor. We also designed an algorithm to find optimal locations to insert synthetic gene networks into the cell so that interference is minimized and orthogonality is maximized. Finally, we developed a method for encrypting and decrypting these genetic 'programs' for the protection of intellectual property. Together these projects explore ways of minimizing...
Multiple mice tracking using Microsoft Kinect - Wang, Chun-Kai, M. Eng. Massachusetts Institute of Technology
Mouse tracking is integral to any attempt to automate mouse behavioral analysis in neuroscience. Systems that rely on vision have successfully tracked a single mouse in one cage, but when attempting to track multiple mice, video-based systems often struggle when the mice interact physically. In this thesis, I develop a novel vision-based tracking system that addresses the challenge of tracking multiple deformable mice with identical appearance, especially during complex occlusions. The system integrates both image and depth modalities to identify the boundary of two occluding mice, and then performs pose estimation to locate nose and tail locations of each mouse....
Risk allocation for temporal risk assessment - Wang, Andrew J
Temporal uncertainty arises when performing any activity in the natural world. When activities are composed into temporal plans, then, there is a risk of not meeting the plan requirements. Currently, we do not have quantitatively precise methods for assessing temporal risk of a plan. Existing methods that deal with temporal uncertainty either forgo probabilistic models or try to optimize a single objective, rather than satisfy multiple objectives. This thesis offers a method for evaluating whether a schedule exists that meets a set of temporal constraints, with acceptable risk of failure. Our key insight is to assume a form of risk...
Design, construction and analysis of an LED strobe controller - Villalva, Nicholas E
In this thesis, I describe an LED strobe controller that I designed and built. With the advent of much brighter LEDs, it is possible to create stroboscopes with both high intensity output and high maximum flash rates. Additionally, the proliferation of Wi-Fi enabled devices provokes the idea of a next-generation interface for tools in Strobe Lab. The initial goal was to produce an LED-based strobe that has sufficient light to replace existing flash tube stroboscopes. The final product combines three high powered LEDs with a microcontroller and Wi-Fi card. Users can adjust settings through a web based interface accessible from...
Embedded avionics with Kalman state estimation for a novel micro-scale unmanned aerial vehicle - Tzanetos, Theodore
An inertial navigation system leveraging Kalman estimation techniques and quaternion dynamics is developed for deployment to a micro-scale unmanned aerial vehicle (UAV). The capabilities, limitations, and requirements of existing navigation solutions motivate the need for an integrated solution that can be readily applied to small embedded systems and still provide reasonably accurate results. Methods to calibrate and compensate systemic inaccuracies in microelectromechanical systems (MEMS) sensors, commonly used in micro-scale UAV applications, are also developed. The problems associated with attitude determination and system localization are analyzed in isolation with incremental simulation and field testing. Performance is evaluated against commercially available inertial...
A system for procedurally generating puzzles for games - Thomson, Alec (Alec P.)
In this thesis, I designed and implemented a set of designer-targeted tools and libraries for procedurally generating puzzles for video games and interactive fiction. The system has the goals of producing solvable and flexible general-purpose puzzles through the use of simple tools targeted at small development teams. The system is implemented as two graphical tools built with the Java Swing Toolkit and a runtime library written for the popular Unity game engine. Two games have been built with earlier versions of the tools and one game was built with the most recent iteration of the tools to test the effectiveness...
A standalone capacitively coupled occupancy sensor - Thompson, William H., M. Eng. Massachusetts Institute of Technology
This thesis presents the design and implementation of a standalone, capacitively coupled, occupancy sensor. Unlike previous iterations, the new sensor is decoupled from the fluorescent lamp. A well controlled, high voltage amplifier and autotransformer are designed as a replacement source and operate over a wide range of output frequencies and amplitudes. The sensor electronics are implemented with active sensing electrodes and resistive and capacitive feedback modes are evaluated for performance. Optimal front end feedback, electrode spacing, and signal source amplitude are explored. The sensor achieves a detection range of 11 feet for occupancy detection, with capacitive measurements down to the...
Development and performance-testing of Multi-Path I/O algorithms on V-Series systems - TerBush, Ryan (Ryan T.)
As data growth continues to accelerate, so must performance and efficiency of large scale storage systems. This project will present the implementation and performance analysis of Multi-Path I/O within Data ONTAP. The goal of this feature is to take advantage of redundant paths that were previously utilized only in failure situations. The paper will address the core mechanisms that comprise the MPIO handling within the system. Furthermore it will present the difficulties of testing such a feature in a shared lab environment. The initial expectation that MPIO would provide a small performance gain, in addition to better failure handling properties,...
Teaching computer science principles using StarLogoTNG - Tannis, Tamika P
This thesis outlines the development of a 3-module set of lesson plans implemented using StarLogoTNG. The purpose of these lesson plans are to serve as a vehicle for teaching and reinforcing specific learning objectives of the CollegeBoard's Advanced Placement Computer Science Principles course, which has 7 main themes. Each lesson plan has as its focus a subset of learning objectives from one of the themes of Creativity, Data, or Internet, while simultaneously incorporating additional learning goals from the themes of Abstraction, Programming, Algorithms, and Impact. These interactive lesson plans go beyond the use of StarLogoTNG to complete specific tasks by...