Class Of Novel Algorithms Research Articles

The Best of Many Worlds: Dual Mirror Descent for Online Allocation Problems

A Novel Class of Robust and Fast Algorithms for Online Allocation Problems A central problem in operations research is allocating limited resources sequentially to maximize cumulative rewards. Applications abound and include network revenue management and internet advertising among many others. Existing data-driven algorithms are tailored for convex settings with either adversarial or stochastic inputs. Many modern applications of online allocations problems, however, are nonconvex. Furthermore, algorithms for adversarial inputs may be too conservative in practice, whereas algorithms for stochastic inputs can perform poorly when the model is misspecified. In the paper “The Best of Many Worlds: Dual Mirror Descent for Online Allocation Problems,” Balseiro, Lu, and Mirrokni present a novel class of algorithms for nonconvex online allocation problems that attain good performance simultaneously in stochastic and adversarial input models and also in various nonstationary settings. The resulting algorithms are simple, fast, and robust to noise and corruption in the observations, in contrast to existing methods from the literature.

BFF and cellhashR: analysis tools for accurate demultiplexing of cell hashing data.

Single-cell sequencing methods provide previously impossible resolution into the transcriptome of individual cells. Cell hashing reduces single-cell sequencing costs by increasing capacity on droplet-based platforms. Cell hashing methods rely on demultiplexing algorithms to accurately classify droplets; however, assumptions underlying these algorithms limit accuracy of demultiplexing, ultimately impacting the quality of single-cell sequencing analyses. We present Bimodal Flexible Fitting (BFF) demultiplexing algorithms BFFcluster and BFFraw, a novel class of algorithms that rely on the single inviolable assumption that barcode count distributions are bimodal. We integrated these and other algorithms into cellhashR, a new R package that provides integrated QC and a single command to execute and compare multiple demultiplexing algorithms. We demonstrate that BFFcluster demultiplexing is both tunable and insensitive to issues with poorly behaved data that can confound other algorithms. Using two well-characterized reference datasets, we demonstrate that demultiplexing with BFF algorithms is accurate and consistent for both well-behaved and poorly behaved input data. cellhashR is available as an R package at https://github.com/BimberLab/cellhashR. cellhashR version 1.0.3 was used for the analyses in this manuscript and is archived on Zenodo at https://www.doi.org/10.5281/zenodo.6402477. Supplementary data are available at Bioinformatics online.

Tensor-Free Proximal Methods for Lifted Bilinear/Quadratic Inverse Problems with Applications to Phase Retrieval

We propose and study a class of novel algorithms that aim at solving bilinear and quadratic inverse problems. Using a convex relaxation based on tensorial lifting, and applying first-order proximal algorithms, these problems could be solved numerically by singular value thresholding methods. However, a direct realization of these algorithms for, e.g., image recovery problems is often impracticable, since computations have to be performed on the tensor-product space, whose dimension is usually tremendous. To overcome this limitation, we derive tensor-free versions of common singular value thresholding methods by exploiting low-rank representations and incorporating an augmented Lanczos process. Using a novel reweighting technique, we further improve the convergence behavior and rank evolution of the iterative algorithms. Applying the method to the two-dimensional masked Fourier phase retrieval problem, we obtain an efficient recovery method. Moreover, the tensor-free algorithms are flexible enough to incorporate a priori smoothness constraints that greatly improve the recovery results.

Pushing the limit of quantum transport simulations

Simulations of quantum transport in coherent conductors have evolved into mature techniques that are used in fields of physics ranging from electrical engineering to quantum nanoelectronics and material science. The most efficient general-purpose algorithms have a computational cost that scales as $L^{6 \dots 7}$ in 3D, which on the one hand is a substantial improvement over older algorithms, but on the other hand still severely restricts the size of the simulation domain, limiting the usefulness of simulations through strong finite-size effects. Here, we present a novel class of algorithms that, for certain systems, allows to directly access the thermodynamic limit. Our approach, based on the Green's function formalism for discrete models, targets systems which are mostly invariant by translation, i.e. invariant by translation up to a finite number of orbitals and/or quasi-1D electrodes and/or the presence of edges or surfaces. Our approach is based on an automatic calculation of the poles and residues of series expansions of the Green's function in momentum space. This expansion allows to integrate analytically in one momentum variable. We illustrate our algorithms with several applications: devices with graphene electrodes that consist of half an infinite sheet; Friedel oscillation calculations of infinite 2D systems in presence of an impurity; quantum spin Hall physics in presence of an edge; the surface of a Weyl semi-metal in presence of impurities and electrodes connected to the surface. In this last example, we study the conduction through the Fermi arcs of the topological material and its resilience to the presence of disorder. Our approach provides a practical route for simulating 3D bulk systems or surfaces as well as other setups that have so far remained elusive.

Efficient MCMC for Gibbs random fields using pre-computation

Bayesian inference of Gibbs random fields (GRFs) is often referred to as a doubly intractable problem, since the normalizing constant of both the likelihood function and the posterior distribution are not in closed-form. The exploration of the posterior distribution of such models is typically carried out with a sophisticated Markov chain Monte Carlo (MCMC) method, the exchange algorithm [28], which requires simulations from the likelihood function at each iteration. The purpose of this paper is to consider an approach to dramatically reduce this computational overhead. To this end we introduce a novel class of algorithms which use realizations of the GRF model, simulated offline, at locations specified by a grid that spans the parameter space. This strategy speeds up dramatically the posterior inference, as illustrated on several examples. However, using the pre-computed graphs introduces a noise in the MCMC algorithm, which is no longer exact. We study the theoretical behaviour of the resulting approximate MCMC algorithm and derive convergence bounds using a recent theoretical development on approximate MCMC methods.

Load Balancing in Cloud-based Flow Security System

In the context of load balancing [1], Lu et al. proposed a novel class of algorithms called Join-Idle-Queue (JIQ) for distributed load balancing in large systems, where a group of dispatchers distribute jobs to a cluster of parallel servers. Each dispatcher maintains a queue called Idle-Queue for idle servers. When a job arrives to a dispatcher at random, the dispatcher sends it to a server which in the head of the Idle-Queue, or to a random server if the Idle-Queue is empty. Meanwhile conversely, when a server has no jobs, it actively requests to be placed on an Idle-Queue. Although this algorithm was shown to be quite effective [2,3], it is defective in some respects. This paper presents several new improvements, which makes the algorithm more effective, especially in Cloud-based Flow Security System (CFS). These improvements focus primarily on these aspects, such as the definition of idle server, data concurrency while choosing an Idle-Queue, etc. The experimental results show that the improved load balancing algorithm can improve the utilization of resources in a certain extent.

Heuristically-Accelerated Multiagent Reinforcement Learning

This paper presents a novel class of algorithms, called Heuristically-Accelerated Multiagent Reinforcement Learning (HAMRL), which allows the use of heuristics to speed up well-known multiagent reinforcement learning (RL) algorithms such as the Minimax-Q. Such HAMRL algorithms are characterized by a heuristic function, which suggests the selection of particular actions over others. This function represents an initial action selection policy, which can be handcrafted, extracted from previous experience in distinct domains, or learnt from observation. To validate the proposal, a thorough theoretical analysis proving the convergence of four algorithms from the HAMRL class (HAMMQ, HAMQ(λ), HAMQS, and HAMS) is presented. In addition, a comprehensive systematical evaluation was conducted in two distinct adversarial domains. The results show that even the most straightforward heuristics can produce virtually optimal action selection policies in much fewer episodes, significantly improving the performance of the HAMRL over vanilla RL algorithms.

Fourth-Order Statistics for Blind Classification of Spatial Multiplexing and Alamouti Space-Time Block Code Signals

Blind signal classification, a major task of intelligent receivers, has important civilian and military applications. This problem becomes more challenging in multi-antenna scenarios due to the diverse transmission schemes that can be employed, e.g., spatial multiplexing (SM) and space-time block codes (STBCs). This paper presents a class of novel algorithms for blind classification of SM and Alamouti STBC (AL-STBC) transmissions. Unlike the prior art, we show that signal classification can be performed using a single receive antenna by taking advantage of the space-time redundancy. The first proposed algorithm relies on the fourth-order moment as a discriminating feature and employs the likelihood ratio test for achieving maximum average probability of correct classification. This requires knowledge of the channel coefficients, modulation type, and noise power. To avoid this drawback, three algorithms have been further developed. Their common idea is that the discrete Fourier transform of the fourth-order lag product exhibits peaks at certain frequencies for the AL-STBC signals, but not for the SM signals, and thus, provides the basis of a useful discriminating feature for signal classification. The effectiveness of these algorithms has been demonstrated in extensive simulation experiments, where a Nakagami-m fading channel and the presence of timing and frequency offsets are assumed.

Join-Idle-Queue: A novel load balancing algorithm for dynamically scalable web services

The prevalence of dynamic-content web services, exemplified by search and online social networking, has motivated an increasingly wide web-facing front end. Horizontal scaling in the Cloud is favored for its elasticity, and distributed design of load balancers is highly desirable. Existing algorithms with a centralized design, such as Join-the-Shortest-Queue ( JSQ), incur high communication overhead for distributed dispatchers. We propose a novel class of algorithms called Join-Idle-Queue ( JIQ) for distributed load balancing in large systems. Unlike algorithms such as Power-of-Two, the JIQ algorithm incurs no communication overhead between the dispatchers and processors at job arrivals. We analyze the JIQ algorithm in the large system limit and find that it effectively results in a reduced system load, which produces 30-fold reduction in queueing overhead compared to Power-of-Two at medium to high load. An extension of the basic JIQ algorithm deals with very high loads using only local information of server load.

A Novel Class of Algorithms for Timing Acquisition of Differential Transmitted Reference UWB Receivers: Architecture, Performance Analysis and System Design

This paper deals with the analysis of the timing acquisition (synchronization) process in Impulse Radio (IR-) Ultra Wide Band (UWB) systems based on Differential Transmitted Reference (DTR) receiver architectures. The fundamental contribution includes a theoretical framework for timing acquisition analysis, which is not only able to describe the first moment (Mean Acquisition Time, MAT), and the second moment (standard deviation of the acquisition time) of the acquisition process, but which also provides a consistent and comprehensive analysis of the probability mass function of the acquisition time, i.e., Acquisition Time Mass Probability. In particular, the specific novelty insights of this paper are as follows: (i) the proposal of a new synchronization algorithm, which is based on a two-step approach and relies on the specific characteristics and peculiarities of DTR receivers, as well as the analysis of its performance via a framework based on the theory of Markov processes; (ii) the exploitation of the LATTICE- POISSON algorithm to accurately estimate the Acquisition Time Mass Probability and Overall Acquisition Probability P <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</sub> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">(ov)</sup> , i.e., the probability to get synchronized before a given time; (iii) the performance analysis and comparison of one-step and two-step solutions with respect to the acquisition threshold; and (iv) the analysis of two approaches to estimate MAT and P <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</sub> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">(ov)</sup> in multipath fading channels, i.e., "Tacq Average" and "Pd Average', which are shown to lead to different results in different scenarios. We observe that restricting the analysis to the MAT may lead to hasty conclusions about the effect of the acquisition threshold on the design of DTR receivers, thus substantiating the need for a more general analysis using P <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</sub> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">(ov)</sup> , as well as the necessity of specifying the characteristics of the wireless channel (e.g., quasi-static, slowly- or fast-fading) for an accurate performance analysis.

Generalized phase retrieval algorithm based on information measures

An iterative phase retrieval algorithm based on the maximum entropy method (MEM) is presented. Introducing a new generalized information measure, we derive a novel class of algorithms which includes the conventionally used error reduction algorithm and a MEM-type iterative algorithm which is presented for the first time. These different phase retrieval methods are unified on the basis of the framework of information measures used in information theory.

Variational optical flow estimation for particle image velocimetry

We introduce a novel class of algorithms for evaluating PIV image pairs. The mathematical basis is a continuous variational formulation for globally estimating the optical flow vector fields over the whole image. This class of approaches has been known in the field of image processing and computer vision for more than two decades but apparently has not been applied to PIV image pairs so far. We pay particular attention to a multi-scale representation of the image data so as to cope with the quite specific signal structure of particle image pairs. The experimental evaluation shows that a prototypical variational approach competes in noisy real-world scenarios with three alternative approaches especially designed for PIV-sequence evaluation. We outline the potential of the variational method for further developments.

Pouzyry: a novel class of algorithms for restoring a function from a random sample

A novel class of algorithms for restoring a function from a random sample is based on the concept of weak convergence, borrows algorithmic solutions from the Optimal Jet Finder (hep-ph/0301185), offers a considerable algorithmic flexibility, is applicable to non-positive functions, is insensitive to the choice of coordinate axes. A first implementation demonstrates feasibility of the approach.

Probabilistic Modeling and Recognition of 3-D Objects

This paper introduces a uniform statistical framework for both 3-D and 2-D object recognition using intensity images as input data. The theoretical part provides a mathematical tool for stochastic modeling. The algorithmic part introduces methods for automatic model generation, localization, and recognition of objects. 2-D images are used for learning the statistical appearance of 3-D objectss both the depth information and the matching between image and model features are missing for model generation. The implied incomplete data estimation problem is solved by the Expectation Maximization algorithm. This leads to a novel class of algorithms for automatic model generation from projections. The estimation of pose parameters corresponds to a non-linear maximum likelihood estimation problem which is solved by a global optimization procedure. Classification is done by the Bayesian decision rule. This work includes the experimental evaluation of the various facets of the presented approach. An empirical evaluation of learning algorithms and the comparison of different pose estimation algorithms show the feasibility of the proposed probabilistic framework.