Search

This paper introduces sTiles, a GPU-accelerated framework for factorizing sparse structured symmetric matrices. By leveraging tile algorithms for fine-grained computations, sTiles uses a structure-aware task execution flow to handle challenging arrowhead sparse matrices with variable bandwidths, common in scientific and engineering fields. It minimizes fill-in during Cholesky factorization using permutation techniques and employs a static scheduler to manage tasks on shared-memory systems with GPU accelerators. sTiles balances tile size and parallelism, where larger tiles enhance algorithmic intensity but increase floating-point operations and memory usage, while parallelism is constrained by the arrowhead structure. To expose more parallelism, a left-looking Cholesky variant breaks sequential dependencies in trailing submatrix updates via tree reductions. Evaluations show sTiles achieves speedups of up to 8.41X, 9.34X, 5.07X, and 11.08X compared to CHOLMOD, SymPACK, MUMPS, and PARDISO, respectively, and a 5X speedup compared to a 32-core AMD EPYC CPU on an NVIDIA A100 GPU. Our generic software framework imports well-established concepts from dense matrix computations but they all require customizations in their deployments on hybrid architectures to best handle factorizations of sparse matrices with arrowhead structures., Comment: 13 pages, 14 figures

Given a convex set $\Omega$ of $\mathbb{R}^n$, we consider the shape optimization problem of finding a convex subset $\omega\subset \Omega$, of a given measure, minimizing the $p$-distance functional $$\mathcal{J}_p(\omega) := \left(\int_{\mathbb{S}^{n-1}} |h_\Omega-h_\omega|^p d\mathcal{H}^{n-1}\right)^{\frac{1}{p}},$$ where $1 \le p <\infty$ and $h_\omega$ and $h_\Omega$ are the support functions of $\omega$ and the fixed container $\Omega$, respectively. We prove the existence of solutions and show that this minimization problem $\Gamma$-converges, when $p$ tends to $+\infty$, towards the problem of finding a convex subset $\omega\subset \Omega$, of a given measure, minimizing the Hausdorff distance to the convex $\Omega$. In the planar case, we show that the free parts of the boundary of the optimal shapes, i.e., those that are in the interior of $\Omega$, are given by polygonal lines. Still in the $2-d$ setting, from a computational perspective, the classical method based on optimizing Fourier coefficients of support functions is not efficient, as it is unable to efficiently capture the presence of segments on the boundary of optimal shapes. We subsequently propose a method combining Fourier analysis and a recent numerical scheme, allowing to obtain accurate results, as demonstrated through numerical experiments.

We develop a method for entangling operations on nuclear spins surrounding a negatively charged boron vacancy (VB-center) point defect in hexagonal boron nitride (hBN). To this end, we propose to employ the electron spin of a VB-center as a control qubit. We show that in the presence of a background magnetic field and by applying control pulses one can collectively manipulate the state of the nuclei with $\hat{U}_z$ and $\hat{U}_x$ rotations. These rotations can serve for implementing the synchronous three-qubit $X$, $Z$, and the Hadamard gates. Through our numerical analyses considering realistic system parameters and the decoherence effects, we demonstrate that these gates can be executed with high fidelities. Furthermore, as an example for the application of our toolbox, we utilize these collective gates to prepare the highly entangled GHZ states among the three nuclear spins with a fidelity of $0.99$. By including the electron decoherence effects we find that the relative deviations of the gate fidelities from the noisy terms are negligibly small, proving the noise-resilience of our protocols. Our work can serve as the groundstone for exploiting the nuclear spins in hBN in future quantum technological applications., Comment: 14 pages, 5 figures

We present a novel approach for accelerating AI performance by leveraging Anderson extrapolation, a vector-to-vector mapping technique based on a window of historical iterations. By identifying the crossover point (Fig. 1) where a mixing penalty is incurred, the method focuses on reducing iterations to convergence, with fewer more compute-intensive but generally cacheable iterations, balancing speed and memory usage with accuracy and algorithmic stability, respectively. We demonstrate significant improvements, in both training and inference, motivated by scalability and efficiency extensions to the realm of high-performance computing (HPC)., Comment: 6 pages, 6 figures, 1 table, Accepted by NeurIPS 2024 Workshop MLNCP https://openreview.net/forum?id=wkP2ZFRn9e

We propose a novel change detection framework to identify changes in the long-term performance behavior of an IaaS service. An IaaS service's long-term performance behavior is represented by an IaaS performance signature. The proposed framework leverages time series similarity measures and a sliding window technique to detect changes in IaaS performance signatures. We introduce a new IaaS performance noise model that enables the proposed framework to distinguish between performance noise and actual changes in performance. The proposed framework utilizes a novel Signal-to-Noise Ratio (SNR) based approach to detect changes when prior knowledge about performance noise is available. A set of experiments is conducted using real-world datasets to demonstrate the effectiveness of the proposed change detection framework., Comment: accepted in ACM transaction on Internet Technology in October 2024. arXiv admin note: text overlap with arXiv:2007.11705

We propose an exercise in which one attempts to deduce the formalism of quantum mechanics from phenomenological observations only. Thus, the only assumed inputs are the multi-time probability distributions estimated from the results of sequential measurements of quantum observables; no presuppositions about the underlying mathematical structures are allowed. We show that it is indeed possible to derive in such a way a complete and fully functional formalism based on the structures of Hilbert spaces. However, the obtained formal description -- the bi-trajectory formalism -- turns out to be quite different from the standard state-focused formalism., Comment: We intend to submit this paper in two parts

Speech signal processing is a cornerstone of modern communication technologies, tasked with improving the clarity and comprehensibility of audio data in noisy environments. The primary challenge in this field is the effective separation and recognition of speech from background noise, crucial for applications ranging from voice-activated assistants to automated transcription services. The quality of speech recognition directly impacts user experience and accessibility in technology-driven communication. This review paper explores advanced clustering techniques, particularly focusing on the Kernel Fuzzy C-Means (KFCM) method, to address these challenges. Our findings indicate that KFCM, compared to traditional methods like K-Means (KM) and Fuzzy C-Means (FCM), provides superior performance in handling non-linear and non-stationary noise conditions in speech signals. The most notable outcome of this review is the adaptability of KFCM to various noisy environments, making it a robust choice for speech enhancement applications. Additionally, the paper identifies gaps in current methodologies, such as the need for more dynamic clustering algorithms that can adapt in real time to changing noise conditions without compromising speech recognition quality. Key contributions include a detailed comparative analysis of current clustering algorithms and suggestions for further integrating hybrid models that combine KFCM with neural networks to enhance speech recognition accuracy. Through this review, we advocate for a shift towards more sophisticated, adaptive clustering techniques that can significantly improve speech enhancement and pave the way for more resilient speech processing systems.

In the noisy intermediate scale quantum (NISQ) era, the control over the qubits is limited due to the errors caused by quantum decoherence, crosstalk, and imperfect calibration. Hence, it is necessary to reduce the size of the large-scale classical data, such as images, when they are to be processed by quantum networks. Conventionally input classical data are reduced in the classical domain using classical networks such as autoencoders and, subsequently, analyzed in the quantum domain. These conventional techniques involve training an enormous number of parameters, making them computationally costly. In this paper, a dynamic patch-based quantum domain data reduction network with a classical attention mechanism is proposed to avoid such data reductions, and subsequently coupled with a novel quantum classifier to perform classification tasks. The architecture processes the classical data sequentially in patches and reduces them using a quantum convolutional-inspired reduction network and further enriches them using a self-attention technique, which utilizes a classical mask derived from simple statistical operations on the native classical data, after measurement. The reduced representation is passed through a quantum classifier, which re-encodes it into quantum states, processes them through quantum ansatzes, and finally measures them to predict classes. This training process involves a joint optimization scheme that considers both the reduction and classifier networks, making the reduction operation dynamic. The proposed architecture has been extensively tested on the publicly available Fashion MNIST dataset, and it has excellent classification performance using significantly reduced training parameters., Comment: 17 pages, 10 figures, 2 tables

The Gottesman-Kitaev-Preskill (GKP) coding is proven to be a good candidate for encoding a qubit on continuous variables (CV) since it is robust under random-shift disturbance. Its preparation in optical systems, however, is challenging to realize in nowadays state-of-the-art experiments. In this article, we propose a simple optical setup for preparing the approximate GKP states by employing a random walk mechanism. We demonstrate this idea by considering the encoding on the transverse position of a single-mode pulse laser. We also discuss generalization and translation to other types of physical CV systems., Comment: 8 pages, 2 figures

The recent surge in AI-generated songs presents exciting possibilities and challenges. While these inventions democratize music creation, they also necessitate the ability to distinguish between human-composed and synthetic songs to safeguard artistic integrity and protect human musical artistry. Existing research and datasets in fake song detection only focus on singing voice deepfake detection (SVDD), where the vocals are AI-generated but the instrumental music is sourced from real songs. However, these approaches are inadequate for detecting contemporary end-to-end artificial songs where all components (vocals, music, lyrics, and style) could be AI-generated. Additionally, existing datasets lack music-lyrics diversity, long-duration songs, and open-access fake songs. To address these gaps, we introduce SONICS, a novel dataset for end-to-end Synthetic Song Detection (SSD), comprising over 97k songs (4,751 hours) with over 49k synthetic songs from popular platforms like Suno and Udio. Furthermore, we highlight the importance of modeling long-range temporal dependencies in songs for effective authenticity detection, an aspect entirely overlooked in existing methods. To utilize long-range patterns, we introduce SpecTTTra, a novel architecture that significantly improves time and memory efficiency over conventional CNN and Transformer-based models. In particular, for long audio samples, our top-performing variant outperforms ViT by 8% F1 score while being 38% faster and using 26% less memory. Additionally, in comparison with ConvNeXt, our model achieves 1% gain in F1 score with 20% boost in speed and 67% reduction in memory usage. Other variants of our model family provide even better speed and memory efficiency with competitive performance., Comment: Updated with correction

Deep AndersoNN accelerates AI by exploiting the continuum limit as the number of explicit layers in a neural network approaches infinity and can be taken as a single implicit layer, known as a deep equilibrium model. Solving for deep equilibrium model parameters reduces to a nonlinear fixed point iteration problem, enabling the use of vector-to-vector iterative solvers and windowing techniques, such as Anderson extrapolation, for accelerating convergence to the fixed point deep equilibrium. Here we show that Deep AndersoNN achieves up to an order of magnitude of speed-up in training and inference. The method is demonstrated on density functional theory results for industrial applications by constructing artificial life and materials `scientists' capable of classifying drugs as strongly or weakly polar, metal-organic frameworks by pore size, and crystalline materials as metals, semiconductors, and insulators, using graph images of node-neighbor representations transformed from atom-bond networks. Results exhibit accuracy up to 98\% and showcase synergy between Deep AndersoNN and machine learning capabilities of modern computing architectures, such as GPUs, for accelerated computational life and materials science by quickly identifying structure-property relationships. This paves the way for saving up to 90\% of compute required for AI, reducing its carbon footprint by up to 60 gigatons per year by 2030, and scaling above memory limits of explicit neural networks in life and materials science, and beyond., Comment: 7 pages, 5 figures, 2 tables, Accepted by ICML ML4LMS https://openreview.net/forum?id=qhwyvhqAvI . International Conference on Machine Learning (ICML). Machine Learning for Life and Material Science (ML4LMS) Workshop, May 2024

Spatial mode demultiplexing was proved to be a successful tool for estimation of the separation between incoherent sources, allowing for sensitivity much below the Rayleigh limit. However, with the presence of measurement's noise, superresolution brought by this technique deteriorates rapidly. On a formal ground, this can be seen in terms of, so called, Rayleigh curse known from direct imaging, which while being absent for ideal spatial mode demultiplexing, goes back in a noisy scenario. In this article, we develop a formal procedure to suppress the destructive effect of the noise, proposing a procedure effectively working as an error correction. For noise models given by a random unitary channel generated by a polynomial of creation and annihilation operators, we demonstrate that perfect noise decoupling can be reached by repeating the mode demultiplexers and intervening them by a group of rotations, in the limit of a large number of repetitions and small noise strength. For a special case of displacement noise, our solution is simplified: by using the demultiplexer twice, and interlacing it by a parity operator, given that the noise configuration is frozen between the first and the second step, a perfect decoupling can be achieved. This allows for a recovery of superresolution for a special class of noise generated by displacement operators. Furthermore, for a strong noise correlation between these two steps, our protocol provides an improved measurement resolution.

The identification and restoration of ancient watermarks have long been a major topic in codicology and history. Classifying historical documents based on watermarks is challenging due to their diversity, noisy samples, multiple representation modes, and minor distinctions between classes and intra-class variations. This paper proposes a modified U-net-based conditional generative adversarial network (GAN) named Npix2Cpix to translate noisy raw historical watermarked images into clean, handwriting-free watermarked images by performing image translation from degraded (noisy) pixels to clean pixels. Using image-to-image translation and adversarial learning, the network creates clutter-free images for watermark restoration and categorization. The generator and discriminator of the proposed GAN are trained using two separate loss functions, each based on the distance between images, to learn the mapping from the input noisy image to the output clean image. After using the proposed GAN to pre-process noisy watermarked images, Siamese-based one-shot learning is employed for watermark classification. Experimental results on a large-scale historical watermark dataset demonstrate that cleaning the noisy watermarked images can help to achieve high one-shot classification accuracy. The qualitative and quantitative evaluation of the retrieved watermarked image highlights the effectiveness of the proposed approach.

Since internet technologies have advanced, one of the primary factors in company development is customer happiness. Online platforms have become prominent places for sharing reviews. Twitter is one of these platforms where customers frequently post their thoughts. Reviews of flights on these platforms have become a concern for the airline business. A positive review can help the company grow, while a negative one can quickly ruin its revenue and reputation. So it's vital for airline businesses to examine the feedback and experiences of their customers and enhance their services to remain competitive. But studying thousands of tweets and analyzing them to find the satisfaction of the customer is quite a difficult task. This tedious process can be made easier by using a machine learning approach to analyze tweets to determine client satisfaction levels. Some work has already been done on this strategy to automate the procedure using machine learning and deep learning techniques. However, they are all purely concerned with assessing the text's sentiment. In addition to the text, the tweet also includes the time, location, username, airline name, and so on. This additional information can be crucial for improving the model's outcome. To provide a machine learning based solution, this work has broadened its perspective to include these qualities. And it has come as no surprise that the additional features beyond text sentiment analysis produce better outcomes in machine learning based models.

The multitime probability distributions obtained by repeatedly probing a quantum system via the measurement of an observable generally violate Kolmogorov's consistency property. Therefore, one cannot interpret such distributions as the result of the sampling of a single trajectory. We show that, nonetheless, they do result from the sampling of one pair of trajectories. In this sense, rather than give up on trajectories, quantum mechanics requires to double down on them. To this purpose, we prove a generalization of the Kolmogorov extension theorem that applies to families of complex-valued bi-probability distributions (that is, defined on pairs of elements of the original sample spaces), and we employ this result in the quantum mechanical scenario. We also discuss the relation of our results with the quantum comb formalism., Comment: Accepted in Quantum

We present a theoretical account of some of the most likely hydrogen-related reactions with impurities in n-type and p-type solar-grade silicon. These include reactions with dopants and carbon, which are relevant in the context of life-time degradation of silicon solar cells, most notably of light and elevated temperature degradation (LeTID) of the cells. Among the problems addressed, we highlight a comparative study of acceptor-enhanced dissociation of hydrogen molecules in B- and Ga-doped material, their subsequent reaction steps toward formation of acceptor-hydrogen pairs, the proposal of mechanisms which explain the observed kinetics of photo-/carrier-induced dissociation of PH and CH pairs in n-type Si, analysis of reactions involving direct interactions between molecules with P and C, and the assignment of several electron and hole traps with detailed atomistic- and wavefunction-resolved models.