Information-Theoretic Confidence Bounds for Reinforcement Learning. (arXiv:1911.09724v1 [stat.ML])

We integrate information-theoretic concepts into the design and analysis of optimistic algorithms and Thompson sampling. By making a connection between information-theoretic quantities and confidence bounds, we obtain results that relate the per-period performance of the agent with its information gain about the environment, thus explicitly characterizing the exploration-exploitation tradeoff. The resulting cumulative regret bound depends…

This Car is Mine!: Automobile Theft Countermeasure Leveraging Driver Identification with Generative Adversarial Networks. (arXiv:1911.09870v1 [cs.CR])

As a car becomes more connected, a countermeasure against automobile theft has become a significant task in the real world. To respond to automobile theft, data mining, biometrics, and additional authentication methods are proposed. Among current countermeasures, data mining method is one of the efficient ways to capture the owner driver’s unique characteristics. To identify…

On the Robustness of Signal Characteristic-Based Sender Identification. (arXiv:1911.09881v1 [cs.CR])

Vehicles become more vulnerable to remote attackers in modern days due to their increasing connectivity and range of functionality. Such increased attack vectors enable adversaries to access a vehicle Electronic Control Unit (ECU). As of today in-vehicle access can cause drastic consequences, because the most commonly used in-vehicle bus technology, the Controller Area Network (CAN),…

UAV-enabled Secure Communication with Finite Blocklength. (arXiv:1911.09887v1 [eess.SP])

In the finite blocklength scenario, which is suitable for practical applications, a method of maximizing the average effective secrecy rate (AESR), significantly distinct from the infinite case, is proposed to optimize the UAV’s trajectory and transmit power subject to UAV’s mobility constraints and transmit power constraints. To address the formulated non-convex optimization problem, it is…

On comparison of estimators for proportional error nonlinear regression models in the limit of small measurement error. (arXiv:1911.09680v1 [math.ST])

In this paper, we compare maximum likelihood (ML), quasi likelihood (QL) and weighted least squares (WLS) estimators for proportional error nonlinear regression models. Literature on thermoluminescence sedimentary dating revealed another estimator similar to weighted least squares but observed responses used as weights. This estimator that we refer to as data weighted least squares (DWLS) is…

Too Quiet in the Library: A Study of Native Third-Party Libraries in Android. (arXiv:1911.09716v1 [cs.CR])

Android applications (“apps”) make avid use of third-party native libraries to increase performance and to reuse already implemented functionality. Native code can be directly executed from apps through the Java Native Interface or the Android Native Development Kit. Android developers drop precompiled native libraries into their projects, enabling their use. Unfortunately, developers are often not…

Communication-Efficient and Byzantine-Robust Distributed Learning. (arXiv:1911.09721v1 [cs.LG])

We develop a communication-efficient distributed learning algorithm that is robust against Byzantine worker machines. We propose and analyze a distributed gradient-descent algorithm that performs a simple thresholding based on gradient norms to mitigate Byzantine failures. We show the (statistical) error-rate of our algorithm matches that of [YCKB18], which uses more complicated schemes (like coordinate-wise median…

EvAn: Neuromorphic Event-based Anomaly Detection. (arXiv:1911.09722v1 [stat.ML])

Event-based cameras are bio-inspired novel sensors that asynchronously record changes in illumination in the form of events, thus resulting in significant advantages over conventional cameras in terms of low power utilization, high dynamic range, and no motion blur. Moreover, such cameras, by design, encode only the relative motion between the scene and the sensor (and…

Fast Sparse ConvNets. (arXiv:1911.09723v1 [cs.CV])

Historically, the pursuit of efficient inference has been one of the driving forces behind research into new deep learning architectures and building blocks. Some recent examples include: the squeeze-and-excitation module, depthwise separable convolutions in Xception, and the inverted bottleneck in MobileNet v2. Notably, in all of these cases, the resulting building blocks enabled not only…

Information-Theoretic Confidence Bounds for Reinforcement Learning. (arXiv:1911.09724v1 [stat.ML])

We integrate information-theoretic concepts into the design and analysis of optimistic algorithms and Thompson sampling. By making a connection between information-theoretic quantities and confidence bounds, we obtain results that relate the per-period performance of the agent with its information gain about the environment, thus explicitly characterizing the exploration-exploitation tradeoff. The resulting cumulative regret bound depends…