Results 1 
8 of
8
Darwin Phones: the Evolution of Sensing and Inference on Mobile Phones
"... We present Darwin, an enabling technology for mobile phone sensing that combines collaborative sensing and classification techniques to reason about human behavior and context on mobile phones. Darwin advances mobile phone sensing through the deployment of efficient but sophisticated machine learnin ..."
Abstract

Cited by 27 (3 self)
 Add to MetaCart
(Show Context)
We present Darwin, an enabling technology for mobile phone sensing that combines collaborative sensing and classification techniques to reason about human behavior and context on mobile phones. Darwin advances mobile phone sensing through the deployment of efficient but sophisticated machine learning techniques specifically designed to run directly on sensorenabled mobile phones (i.e., smartphones). Darwin tackles three key sensing and inference challenges that are barriers to massscale adoption of mobile phone sensing applications: (i) the humanburden of training classifiers, (ii) the ability to perform reliably in different environments (e.g., indoor, outdoor) and (iii) the ability to scale to a large number of phones without jeopardizing the “phone experience ” (e.g., usability and battery lifetime). Darwin is a collaborative reasoning framework built on three concepts: classifier/model evolution, model pooling, and collaborative inference. To the best of our knowledge Darwin is the first system that applies distributed machine learning techniques and collaborative inference concepts to mobile phones. We implement the Darwin system on the Nokia N97 and Apple iPhone. While Darwin represents a general framework applicable to a wide variety of emerging mobile sensing applications, we implement a speaker recognition application and an augmented reality application to evaluate the benefits of Darwin. We show experimental results from eight individuals carrying Nokia N97s and demonstrate that Darwin improves the reliability and scalability of the proofofconcept speaker recognition application without additional burden to users.
Differentially private iterative synchronous consensus
 in Proceedings of the CCS Workshop on Privacy in the Electronic Society (WPES
, 2012
"... The iterative consensus problem requires a set of processes or agents with different initial values, to interact and update their states to eventually converge to a common value. Protocols solving iterative consensus serve as building blocks in a variety of systems where distributed coordination i ..."
Abstract

Cited by 5 (1 self)
 Add to MetaCart
(Show Context)
The iterative consensus problem requires a set of processes or agents with different initial values, to interact and update their states to eventually converge to a common value. Protocols solving iterative consensus serve as building blocks in a variety of systems where distributed coordination is required for load balancing, data aggregation, sensor fusion, filtering, clock synchronization and platooning of autonomous vehicles. In this paper, we introduce the private iterative consensus problem where agents are required to converge while protecting the privacy of their initial values from honest but curious adversaries. Protecting the initial states, in many applications, suffice to protect all subsequent states of the individual participants. First, we adapt the notion of differential privacy in this setting of iterative computation. Next, we present a serverbased and a completely distributed randomized mechanism for solving private iterative consensus with adversaries who can observe the messages as well as the internal states of the server and a subset of the clients. Finally, we establish the tradeoff between privacy and the accuracy of the proposed randomized mechanism. 1
Reaching Consensus with Uncertainty on a Network
, 2009
"... As modern communication networks become increasingly advanced, so does the ability and necessity to communicate to make more informed decisions. However, communication alone is not sufficient; the method by which information is incorporated and used to make the decision is of critical importance. Th ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
As modern communication networks become increasingly advanced, so does the ability and necessity to communicate to make more informed decisions. However, communication alone is not sufficient; the method by which information is incorporated and used to make the decision is of critical importance. This thesis develops a novel distributed agreement protocol that allows multiple agents to agree upon a parameter vector particularly when each agent has a unique measure of possibly nonGaussian uncertainty in its estimate. The proposed hyperparameter consensus algorithm builds upon foundations in both the consensus and data fusion communities by applying Bayesian probability theory to the agreement problem. Unique to this approach is the ability to converge to the centralized Bayesian parameter estimate of nonGaussian distributed variables over arbitrary, strongly connected networks and without the burden of the often prohibitively complex filters used in traditional data fusion solutions. Convergence properties are demonstrated for local estimates described by a number of common probability distributions and over a
Decentralized Kalman Filter Comparison for DistributedParameter Systems: A Case Study for a 1D Heat Conduction Process
"... comparison for distributedparameter systems: A case study for a 1D heat conduction process ∗ ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
comparison for distributedparameter systems: A case study for a 1D heat conduction process ∗
Dean of Graduate Studiesc ○ 2011
, 2011
"... The increasing popularity of smartphones with their embedded sensing capability and the availability of new application distribution channels, such as, the Apple AppStore and the Google Android Market, is giving researchers a unique opportunity to deploy mobile sensing applications at unprecedented ..."
Abstract
 Add to MetaCart
(Show Context)
The increasing popularity of smartphones with their embedded sensing capability and the availability of new application distribution channels, such as, the Apple AppStore and the Google Android Market, is giving researchers a unique opportunity to deploy mobile sensing applications at unprecedented scale and collect sensor data way beyond the boundaries of traditional smallscale research laboratory deployments. This thesis makes a number of contributions to smartphone sensing by introducing new sensing models, algorithms, applications, and systems. First, we propose CenceMe, the first largescale personal and social sensing application for smartphones, which allows users to share their realtime “sensing presence ” (i.e., activity and context) with friends using the phone, web, and social network sites (i.e., Facebook, Myspace, Twitter). CenceMe exploits the smartphone’s onboard sensors (viz. accelerometer, microphone, GPS, Bluetooth, WiFi, camera) and lightweight, efficient machine learning algorithms on the phone and backend servers to automatically infer people’s activity and social context (e.g., having a conversation, in a meeting, at a party). The development, deployment, and evaluation of CenceMe opened up new problems also studied in this dissertation. Sensing with smartphones presents several technical challenges that need to be surmounted; for
Algebraic Connectivity Ratio of Ramanujan Graphs
"... Abstract — In this paper, we explore spectral properties of a class of regular Cayley graphs known as Ramanujan graphs and prove that the ratio of their algebraic connectivity to that of regular lattices grows exponentially as O(n γ) with γ = 1.84±0.05 for networks with average degree of O(log(n)). ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract — In this paper, we explore spectral properties of a class of regular Cayley graphs known as Ramanujan graphs and prove that the ratio of their algebraic connectivity to that of regular lattices grows exponentially as O(n γ) with γ = 1.84±0.05 for networks with average degree of O(log(n)). Explicit construction algorithms exist for Ramanujan graphs that create regular graphs with especial degree and scale that depend on a pair of prime numbers. We introduce a randomized algorithm for construction of a class of fast regular graphs called quasi Ramanujan graphs. These graphs are obtained from finite number of degree balancing operations on WattsStrogatz smallword networks that are irregular graphs. We show that quasi Ramanujan graphs share similar combinatorial optimality spectral properties as Ramanujan graphs and are not restricted to especial choices of degree and scale. A byproduct of this fact is that the algebraic connectivity ratio of quasi Ramanujan graphs grows exponentially in n as well. Numerical experiments are performed to verify our analytical predictions. Consensus algorithms converge extremely fast on networks with exponentially growing algebraic connectivity ratios. Index Terms — random graphs, Ramanujan graphs, Cayley graphs, smallworld networks, algebraic connectivity, ultrafast consensus I.
on a Network
, 2009
"... As modern communication networks become increasingly advanced, so does the ability and necessity to communicate to make more informed decisions. However, communication alone is not sufficient; the method by which information is incorporated and used to make the decision is of critical importance. T ..."
Abstract
 Add to MetaCart
(Show Context)
As modern communication networks become increasingly advanced, so does the ability and necessity to communicate to make more informed decisions. However, communication alone is not sufficient; the method by which information is incorporated and used to make the decision is of critical importance. This thesis develops a novel distributed agreement protocol that allows multiple agents to agree upon a parameter vector particularly when each agent has a unique measure of possibly nonGaussian uncertainty in its estimate. The proposed hyperparameter consensus algorithm builds upon foundations in both the consensus and data fusion communities by applying Bayesian probability theory to the agreement problem. Unique to this approach is the ability to converge to the centralized Bayesian parameter estimate of nonGaussian distributed variables over arbitrary, strongly connected networks and without the burden of the often prohibitively complex filters used in traditional data fusion solutions. Convergence properties are demonstrated for local estimates described by a number of common probability distributions and over a