Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Course Description In this course, we will explore the question of the actual and potential connections between democracy and education. Our focus of attention will be placed on a critical examination of democratic theory and its implications for the civic education roles and contributions of teachers, adult educators, community development practitioners, and community organizers. We will survey and deal critically with a range of competing conceptions of democracy, variously described as classical, republican, liberal, radical, marxist, neomarxist, pragmatist, feminist, populist, pluralist, postmodern, and/or participatory. Using narrative inquiry as a means for illuminating and interpreting contemporary practice, we will analyze the implications of different conceptions of democracy for the practical work of civic education.

Democracy and education.

Normal accidents. Living with high-risk technologies

我的台灣, 看見心靈的故鄉 =2009林磐聳藝術與設計展

/pdf/millennials-rising-the-next-great-generation-3011lw78bf.pdf

Millennials Rising: The Next Great Generation

As part of the Detecting Intrusions at Layer ONe (DILON) project, we show that Ethernet devices can be uniquely identified and tracked—using as few as 25 Ethernet frames—by analyzing variations in their analog signal caused by hardware and manufacturing inconsistencies. An optimal detector, the matched filter, is utilized to create signal profiles, which aid in identifying the device the signal originated from. Several non-traditional applications of the filter are presented in order to improve its ability to discriminate between signals from seemingly identical devices of the same manufacturing lot. The experimental results of applying these filters to three different models of Ethernet cards, totaling 16 devices, are presented and discussed. Important applications of this technology include intrusion detection (discovering node impersonation and network tampering), authentication (preventing unauthorized access to the physical network), forensic data collection (tying a physical device to a specific network incident), and assurance monitoring (determining whether a device will or is in the process of failing).

Device Identification via Analog Signal Fingerprinting: A Matched Filter Approach

In this paper, we show that a single, maliciously controlled vehicle can destabilize a vehicular platoon, to catastrophic effect, through local modifications to the prevailing control law. Specifically, by combining changes to the gains of the associated law with the appropriate vehicle movements, the attacker can cause the platoon to oscillate at a resonant frequency, causing accidents that could result in serious injury or death. We determine the range of gains, and their corresponding frequencies, that allow an attacker to violate the string stability and stability criteria at different positions in the platoon. Furthermore, we prove that the attack can be successful at any position in the platoon and at frequencies that can be realized by the other vehicles in the platoon. Our work implies that neither the string stability nor stability conditions, when used singly, ensure proper platoon operation, and that neither can be used to ensure the other. Finally, we show that an attacker is theoretically capable of gaining control over the individual position and velocity (states) of other vehicles in the platoon; two attacks are demonstrated for this vulnerability.

Vehicular Platooning in an Adversarial Environment

Future automated vehicles will rely on V2V communication to exchange information about their motion states and take corresponding control actions, to enhance road safety and efficiency. Evaluating the trustworthiness of such data in a VANET is critical as malicious vehicles may inject false data which will undermine the benefits of V2V communication and lead to severe consequences, such as collisions. Existing solutions are inadequate since they assume an honest majority of vehicles. In this work, we propose a novel data trust framework, which determines the truthfulness of each received message on the fly, and is able to detect false data and securely track vehicles even when they report false information. The basic idea is to verify the implied effect of vehicle's reported data using secure sensing mechanisms from the wireless physical layer, which is wrapped within a dynamic vehicle tracking system using extended Kalman filter. Our framework does require at least one honest neighboring vehicle, and simulation results show that it is effective in most highway traffic scenarios.

A data trust framework for VANETs enabling false data detection and secure vehicle tracking

This work describes a new type of efficiency attack that can be used to degrade the performance of automated vehicular transportation systems. Next-generation transportation technologies will leverage increasing use of vehicle automation. Proposed vehicular automation systems include cooperative adaptive cruise control and vehicle platooning strategies which require cooperation and coordination among vehicles. These strategies are intended to optimize through-put and energy usage in future highway systems, but, as we demonstrate, they also introduce new vulnerabilities. In this work we show that a typical platooning system would allow a maliciously controlled vehicle to exert subtle influence on the motion of surrounding vehicles. This effect can be used to increase the energy expenditure of surrounding vehicles by 20% to 300%.

CPS: an efficiency-motivated attack against autonomous vehicular transportation

Specific Emitter Identification (SEI) is the act of matching a received signal to an emitter using a database of radio frequency (RF) features belonging to known transmitters. SEI systems are of vital importance to the military for applications such as early warning systems, emitter tracking, and emitter location, and, more recently, have been used in cognitive radio systems to enforce Dynamic Spectrum Access (DSA) rules [1], [2]. This work investigates using Convolutional Neural Networks (CNNs) as feature learners and extractors, paired with the clustering algorithm DBSCAN, to perform SEI. The process through which emitter-specific features are extracted from raw I/Q data streams is described in detail, including the CNN architecture, design, and training. Extensive performance analysis demonstrates the effectiveness of the proposed approach in identifying emitters, and shows that features extracted from CNNs can be used to differentiate between devices unseen in training.

Ryan M. Gerdes

Papers

Device Identification via Analog Signal Fingerprinting: A Matched Filter Approach

Vehicular Platooning in an Adversarial Environment

A data trust framework for VANETs enabling false data detection and secure vehicle tracking

CPS: an efficiency-motivated attack against autonomous vehicular transportation

Clustering Learned CNN Features from Raw I/Q Data for Emitter Identification