Search

In this work, we take the initiative in studying the information-theoretic tradeoff between communication and quickest change detection (QCD) under an integrated sensing and communication setting. We formally establish a joint communication and sensing problem for the quickest change detection. We assume a broadcast channel with a transmitter, a communication receiver, and a QCD detector in which only the detection channel is state dependent. For the problem setting, by utilizing constant subblock-composition codes and a modified CuSum detection rule, which we call subblock CuSum (SCS), we provide an inner bound on the information-theoretic tradeoff between communication rate and change point detection delay in the asymptotic regime of vanishing false alarm rate. We further provide a partial converse that matches our inner bound for a certain class of codes. This implies that the SCS detection strategy is asymptotically optimal for our codes as the false alarm rate constraint vanishes. We also present some canonical examples of the tradeoff region for a binary channel, a scalar Gaussian channel, and a MIMO Gaussian channel.

Recent advancements in deep learning models have significantly enhanced jet classification performance by analyzing low-level features (LLFs). However, this approach often leads to less interpretable models, emphasizing the need to understand the decision-making process and to identify the high-level features (HLFs) crucial for explaining jet classification. To address this, we consider the top jet tagging problems and introduce an analysis model (AM) that analyzes selected HLFs designed to capture important features of top jets. Our AM mainly consists of the following three modules: a relation network analyzing two-point energy correlations, mathematical morphology and Minkowski functionals for generalizing jet constituent multiplicities, and a recursive neural network analyzing subjet constituent multiplicity to enhance sensitivity to subjet color charges. We demonstrate that our AM achieves performance comparable to the Particle Transformer (ParT) while requiring fewer computational resources in a comparison of top jet tagging using jets simulated at the hadronic calorimeter angular resolution scale. Furthermore, as a more constrained architecture than ParT, the AM exhibits smaller training uncertainties because of the bias-variance tradeoff. We also compare the information content of AM and ParT by decorrelating the features already learned by AM. Lastly, we briefly comment on the results of AM with finer angular resolution inputs., Comment: 36 pages, 8 figures, 12 tables, version published in JHEP. A short discussion on the analysis model and high-level features at sub-hadronic calorimeter angular resolution is added. Code for computing Minkowski functionals is available at: https://github.com/sunghak-lim/mfjet

Machine learning (ML) has recently shown promising results in medical predictions using electronic health records (EHRs). However, since ML models typically have a limited capability in terms of input sizes, selecting specific medical events from EHRs for use as input is necessary. This selection process, often relying on expert opinion, can cause bottlenecks in development. We propose Retrieval-Enhanced Medical prediction model (REMed) to address such challenges. REMed can essentially evaluate unlimited medical events, select the relevant ones, and make predictions. This allows for an unrestricted input size, eliminating the need for manual event selection. We verified these properties through experiments involving 27 clinical prediction tasks across four independent cohorts, where REMed outperformed the baselines. Notably, we found that the preferences of REMed align closely with those of medical experts. We expect our approach to significantly expedite the development of EHR prediction models by minimizing clinicians' need for manual involvement., Comment: The source codes corresponding to this paper are available at: https://github.com/starmpcc/REMed

In many sensor network applications, a fusion center often has additional valuable information, such as context data, which cannot be obtained directly from the sensors. Motivated by this, we study a generalized CEO problem where a CEO has access to context information. The main contribution of this work is twofold. Firstly, we characterize the asymptotically optimal error exponent per rate as the number of sensors and sum rate grow without bound. The proof extends the Berger-Tung coding scheme and the converse argument by Berger et al. (1996) taking into account context information. The resulting expression includes the minimum Chernoff divergence over context information. Secondly, assuming that the sizes of the source and context alphabets are respectively $|\mathcal{X}|$ and $|\mathcal{S}|$, we prove that it is asymptotically optimal to partition all sensors into at most $\binom{|\mathcal{X}|}{2} |\mathcal{S}|$ groups and have the sensors in each group adopt the same encoding scheme. Our problem subsumes the original CEO problem by Berger et al. (1996) as a special case if there is only one letter for context information; in this case, our result tightens its required number of groups from $\binom{|\mathcal{X}|}{2}+2$ to $\binom{|\mathcal{X}|}{2}$. We also numerically demonstrate the effect of context information for a simple Gaussian scenario.

We present a novel, data-driven analysis of Galactic dynamics, using unsupervised machine learning -- in the form of density estimation with normalizing flows -- to learn the underlying phase space distribution of 6 million nearby stars from the Gaia DR3 catalog. Solving the collisionless Boltzmann equation with the assumption of approximate equilibrium, we calculate -- for the first time ever -- a model-free, unbinned, fully 3D map of the local acceleration and mass density fields within a 3 kpc sphere around the Sun. As our approach makes no assumptions about symmetries, we can test for signs of disequilibrium in our results. We find our results are consistent with equilibrium at the 10% level, limited by the current precision of the normalizing flows. After subtracting the known contribution of stars and gas from the calculated mass density, we find clear evidence for dark matter throughout the analyzed volume. Assuming spherical symmetry and averaging mass density measurements, we find a local dark matter density of $0.47\pm 0.05\;\mathrm{GeV/cm}^3$. We fit our results to a generalized NFW, and find a profile broadly consistent with other recent analyses., Comment: 19 pages, 13 figures, 3 tables

Cosmological N-body simulations of galaxies operate at the level of "star particles" with a mass resolution on the scale of thousands of solar masses. Turning these simulations into stellar mock catalogs requires "upsampling" the star particles into individual stars following the same phase-space density. In this paper, we introduce two new upsampling methods. First, we describe GalaxyFlow, a sophisticated upsampling method that utilizes normalizing flows to both estimate the stellar phase space density and sample from it. Second, we improve on existing upsamplers based on adaptive kernel density estimation, using maximum likelihood estimation to fine-tune the bandwidth for such algorithms in a way that improves both the density estimation accuracy and upsampling results. We demonstrate our upsampling techniques on a neighborhood of the Solar location in two simulated galaxies: Auriga 6 and h277. Both yield smooth stellar distributions that closely resemble the stellar densities seen in the Gaia DR3 catalog. Furthermore, we introduce a novel multi-model classifier test to compare the accuracy of different upsampling methods quantitatively. This test confirms that GalaxyFlow estimates the density of the underlying star particles more accurately than methods based on kernel density estimation, at the cost of being more computationally intensive., Comment: 22 pages, 14 figures, version published in MNRAS

Measuring the density profile of dark matter in the Solar neighborhood has important implications for both dark matter theory and experiment. In this work, we apply autoregressive flows to stars from a realistic simulation of a Milky Way-type galaxy to learn -- in an unsupervised way -- the stellar phase space density and its derivatives. With these as inputs, and under the assumption of dynamic equilibrium, the gravitational acceleration field and mass density can be calculated directly from the Boltzmann Equation without the need to assume either cylindrical symmetry or specific functional forms for the galaxy's mass density. We demonstrate our approach can accurately reconstruct the mass density and acceleration profiles of the simulated galaxy, even in the presence of Gaia-like errors in the kinematic measurements., Comment: 23 pages, 9 figures

Deep-learning-based intelligent services have become prevalent in cyber-physical applications including smart cities and health-care. Deploying deep-learning-based intelligence near the end-user enhances privacy protection, responsiveness, and reliability. Resource-constrained end-devices must be carefully managed in order to meet the latency and energy requirements of computationally-intensive deep learning services. Collaborative end-edge-cloud computing for deep learning provides a range of performance and efficiency that can address application requirements through computation offloading. The decision to offload computation is a communication-computation co-optimization problem that varies with both system parameters (e.g., network condition) and workload characteristics (e.g., inputs). On the other hand, deep learning model optimization provides another source of tradeoff between latency and model accuracy. An end-to-end decision-making solution that considers such computation-communication problem is required to synergistically find the optimal offloading policy and model for deep learning services. To this end, we propose a reinforcement-learning-based computation offloading solution that learns optimal offloading policy considering deep learning model selection techniques to minimize response time while providing sufficient accuracy. We demonstrate the effectiveness of our solution for edge devices in an end-edge-cloud system and evaluate with a real-setup implementation using multiple AWS and ARM core configurations. Our solution provides 35% speedup in the average response time compared to the state-of-the-art with less than 0.9% accuracy reduction, demonstrating the promise of our online learning framework for orchestrating DL inference in end-edge-cloud systems.

We propose a hybrid coded modulation scheme which composes of inner and outer codes. The outer-code can be any standard binary linear code with efficient soft decoding capability (e.g. low-density parity-check (LDPC) codes). The inner code is designed using a deep neural network (DNN) which takes the channel coded bits and outputs modulated symbols. For training the DNN, we propose to use a loss function that is inspired by the generalized mutual information. The resulting constellations are shown to outperform the conventional quadrature amplitude modulation (QAM) based coding scheme for modulation order 16 and 64 with 5G standard LDPC codes.

Three-dimensional integrated circuits promise power, performance, and footprint gains compared to their 2D counterparts, thanks to drastic reductions in the interconnects' length through their smaller form factor. We can leverage the potential of 3D integration by enhancing MemPool, an open-source many-core design with 256 cores and a shared pool of L1 scratchpad memory connected with a low-latency interconnect. MemPool's baseline 2D design is severely limited by routing congestion and wire propagation delay, making the design ideal for 3D integration. In architectural terms, we increase MemPool's scratchpad memory capacity beyond the sweet spot for 2D designs, improving performance in a common digital signal processing kernel. We propose a 3D MemPool design that leverages a smart partitioning of the memory resources across two layers to balance the size and utilization of the stacked dies. In this paper, we explore the architectural and the technology parameter spaces by analyzing the power, performance, area, and energy efficiency of MemPool instances in 2D and 3D with 1 MiB, 2 MiB, 4 MiB, and 8 MiB of scratchpad memory in a commercial 28 nm technology node. We observe a performance gain of 9.1% when running a matrix multiplication on the MemPool-3D design with 4 MiB of scratchpad memory compared to the MemPool 2D counterpart. In terms of energy efficiency, we can implement the MemPool-3D instance with 4 MiB of L1 memory on an energy budget 15% smaller than its 2D counterpart, and even 3.7% smaller than the MemPool-2D instance with one-fourth of the L1 scratchpad memory capacity., Comment: Accepted for publication in DATE 2022 -- Design, Automation and Test in Europe Conference

Compute-forward is a coding technique that enables receiver(s) in a network to directly decode one or more linear combinations of the transmitted codewords. Initial efforts focused on Gaussian channels and derived achievable rate regions via nested lattice codes and single-user (lattice) decoding as well as sequential (lattice) decoding. Recently, these results have been generalized to discrete memoryless channels via nested linear codes and joint typicality coding, culminating in a simultaneous-decoding rate region for recovering one or more linear combinations from $K$ users. Using a discretization approach, this paper translates this result into a simultaneous-decoding rate region for a wide class of continuous memoryless channels, including the important special case of Gaussian channels. Additionally, this paper derives a single, unified expression for both discrete and continuous rate regions via an algebraic generalization of R\'enyi's information dimension., Comment: 86 pages, 7 figures, submitted to IEEE Transactions of Information Theory