Coordinate Descent Approach Research Articles

Structured neural networks for CPUE standardization: A case study of the blue endeavour prawn in Australia's Northern Prawn Fishery

We performed a study on standardizing the catch per unit effort (CPUE) for blue endeavour prawns (Metapenaeus endeavouri) caught in the Northern Prawn Fishery (NPF), one of Australia’s largest and most valuable prawn fisheries. Blue endeavour prawns constitute a significant proportion of the total NPF catches. However, there have been very limited studies on their population dynamics. This study assessed the effectiveness of Artificial Neural Networks (ANNs) for CPUE standardization, with a focus on blue endeavour prawns as a case study. Our approach involved developing new ANN models for CPUE standardization with two key ideas: using an architecture inspired by the catch equation to mitigate overfitting; and using the Tweedie distribution to manage uncertainties and zero counts in the catch data. Specifically, we grouped variables into three distinct modules based on the catch equation, with each representing catchability, fishing effort, and fish density, respectively. Parameter estimation for our ANNs was achieved by maximizing the likelihood using a coordinate descent approach, which alternates between optimizing the Tweedie distribution parameters (power and dispersion) and the standard neural net parameters. We conducted a comprehensive comparison among ANNs, generalized linear models, and generalized additive models. The findings suggest that customizing ANN structure improves model fitting and effectively mitigates the risk of overfitting. It also reveals a promising path for the application of neural networks in CPUE standardization.

Joint Transmit and Receive Beamforming Design for DPC-Based MIMO DFRC Systems

This paper proposes an optimal beamforming strategy for a downlink multi-user multi-input–multi-output (MIMO) dual-function radar communication (DFRC) system with dirty paper coding (DPC) adopted at the transmitter. We aim to achieve the maximum weighted sum rate of communicating users while adhering to a predetermined transmit covariance constraint for radar performance assurance. To make the intended problem trackable, we leverage the equivalence of the weighted sum rate and the weighted minimum mean squared error (MMSE) to reframe the issue and devise a block coordinate descent (BCD) approach to iteratively calculate transmit and receive beamforming solutions. Through this methodology, we demonstrate that the optimal receive beamforming aligns with the traditional MMSE approach, whereas the optimal transmit beamforming design can be cast into a quadratic optimization problem defined on a complex Stiefel manifold. Based on the majorization–minimization (MM) method, an iterative algorithm is then developed to compute the optimal transmit beamforming design by solving a series of orthogonal Procrustes problems (OPPs) that admit closed-form optimal solutions. Numerical findings serve to validate the efficacy of our scheme. It is demonstrated that our approach can achieve at least 73% higher spectral efficiency than the existing methods in a high signal-to-noise ratio (SNR) regime.

A Coordinate Descent Approach to Atomic Norm Denoising

A Coordinate Descent Approach to Atomic Norm Denoising

Secure two-way transmission via untrusted UAV Relay: Joint path design and slot-pairing strategy

This paper proposes a novel two-way secure communication system that allows numerous pairs of ground users (GUs) communicate with one another using an untrusted UAV relay (UUR), where the UUR can eavesdrop on information while helping GUs perform two-way information transmission. We propose a safe scheme using slot pairing and joint path design to ensure the security of communication. Specifically, according to the physical-layer network coding (PNC) protocol, the collected information from GUs is cached in the UAV’s shock absorber before being transmitted as the UAV approaches optimal channel conditions. By matching the channel status of two-way transmission with the slot-pairing approach, our goal is to optimize the system’s secrecy sum rate (SSR). Due to the hybrid integer non-convex optimization problem, the formulated problem is difficult to address. As a result, we decompose it into three subproblems and propose an effective iterative algorithm that uses the block coordinate descent (BCD) approach and the sequential convex approximation (SCA) technique. The convergence and efficacy of our proposed scheme are demonstrated by numerical results.

Secure Beamforming in Multi-User Multi-IRS Millimeter Wave Systems

We study the secrecy rate maximization problem in a millimeter wave (mmWave) network, consisting of a base station (BS), multiple intelligent reflecting surfaces (IRSs) (or reconfigurable intelligent surfaces (RISs)), multiple users, and a single eavesdropper. To ensure a fair secrecy rate among all the users, we adopt a max-min fairness criterion which results in a mixed integer problem. We first relax discrete IRSs phase shifts to the continuous ones. To cope with the non-convexity of the relaxed optimization problem, we leverage the penalty method and block coordinate descent approach to divide it into two sub-problems, which are solved by successive convex approximation (SCA). Then, we propose a low-complexity mapping algorithm where feasible IRSs phase shifts are obtained. Mathematical evaluation shows the convergence of sub-problems to a Karush-Kuhn-Tucker (KKT) point of the original ones. Furthermore, the convergence guarantee of the overall proposed algorithm and computational complexity are investigated. Finally, simulation results show our proposed algorithm outweighs the conventional solutions based on the semi-definite programming (SDP) in terms of convergence and secrecy rate, especially in a larger number of IRSs and phase shifts where SDP suffers from rank-one approximation. Maximum ratio transmission (MRT) and IRS-free systems are also considered as other benchmarks.

RBDL: Robust block-Structured dictionary learning for block sparse representation

Dictionary learning methods have been extensively used in different types of image and signal processing tasks. In a number of applications, the collected data/signal may have a multi-subspace structure and be perturbed with outliers. These motivate the use of robust and block-sparse signal representations. In this paper, a new algorithm for learning a block-structured dictionary in the presence of outliers is proposed. It is based on α−divergence and has the advantage of tolerating the presence of outliers. A block coordinate descent approach is adopted to obtain simple closed-form solutions for both the sparse coding and dictionary update stages. Finally, experimental results illustrating the superiority of the proposed method over some state-of-the-art dictionary learning methods, are provided.

Robust Resource Allocation Design for Secure IRS-Aided WPCN

This paper studies the robust resource allocation design for secure intelligent reflecting surface (IRS)-aided wireless powered communication networks (WPCN). In particular, deploying an IRS can establish favorable end-to-end radio propagation environment for achieving the desired performance gain in secure wireless-powered systems. We aim to minimize the total hybrid base station (HBS) transmit power by jointly designing the active transmitting and receiving beamforming at the HBS, the passive beamforming at the IRS, and the transmit power of each wireless-powered device (WD) and jammer node (JN). We formulate a non-convex optimization problem for the robust resource allocation design taking into account the secrecy rate requirement of the WDs and the power budgets for both the WDs and the JNs. To handle this intractable problem, we propose a computationally efficient iterative suboptimal algorithm exploiting the block coordinate descent approach, the successive convex approximation, and the penalty method, which attains a Karush-Kuhn-Tucker (KKT) solution of the transformed problem. Also, we reveal that the optimal energy beamforming matrices are rank-one sharing the same spatial direction. Simulation results unveil that the proposed scheme is able to dramatically reduce the HBS transmit power over various baseline schemes adopting existing solutions. Besides, our results show the superiority of introducing IRS for secure communication in wireless-powered systems.

Numerical methods for CT reconstruction with unknown geometry parameters

Computed tomography (CT) techniques are well known for their ability to produce high-quality images needed for medical diagnostic purposes. Unfortunately, standard CT machines are extremely large, heavy, require careful and regular calibration, and are expensive, which can limit their availability in point-of-care situations. An alternative approach is to use portable machines, but parameters related to the geometry of these devices (e.g., distance between source and detector, orientation of source to detector) cannot always be precisely calibrated, and these parameters may change slightly when the machine is adjusted during the image acquisition process. In this work, we describe the non-linear inverse problem that models this situation, and discuss algorithms that can jointly estimate the geometry parameters and compute a reconstructed image. In particular, we propose a hybrid machine learning and block coordinate descent (ML-BCD) approach that uses an ML model to calibrate geometry parameters, and uses BCD to refine the predicted parameters and reconstruct the imaged object simultaneously. We show using numerical experiments that our new method can efficiently improve the accuracy of both the image and geometry parameters.

Binary multi-modal matrix factorization for fast item cold-start recommendation

Hashing technology can support large-scale recommendation very effectively due to its advantages of low storage cost and high recommendation efficiency. However, existing hashing-based recommendation methods often suffer from item cold-start problem. This is because they simply consider the user-item interaction and the single content information of the items, but the full interaction history is not always available and the single auxiliary information may be missing. To solve this issue, in this paper we propose a Binary Multi-modal Matrix Factorization (BMMF) method. First, we propose an efficient consensus multi-modal mapping to transform the heterogeneous multi-modal features to the unified factors by exploiting the complementarity of multiple modalities. Then, binary matrix factorization is simultaneously performed on the multi-modal features of the items and past user preferences to learn the compact binary codes of the users/items in a common Hamming space. In addition, inspired by the observation that similar instances often have similar binary codes within a short Hamming distance, we formulate a semantic structure regularization term to preserve the similarities of the items during the binary embedding process. Finally, we develop an effective Discrete Coordinate Descent (DCD) approach to tackle the formulated discrete hash optimization problem directly. Experiments on three publicly available real-world datasets demonstrate the superiority of the proposed method against the state-of-the-art methods. Our source codes and testing datasets are available athttps://github.com/pcm1217/BMMF.

Sparse Deconvolution of Pulsatile Growth Hormone Secretion in Adolescents.

Growth hormone (GH) is secreted by cells in the anterior pituitary on two time scales: discrete pulses over minutes that occur within a 24-hr pattern. Secretion reflects the balance of stimulatory and inhibitory inputs from the hypothalamus and is influenced by gonadal steroids, stress, nutrition, and sleep/wake states. We propose a novel approach for the analysis of GH data and use this approach to quantify (i) the timing, amplitude and the number of GH pulses and (ii) GH infusion, clearance and basal secretion (i.e., time invariant) rates, using serum GH sampled every 10 minutes during an 8-hour sleep study in 18 adolescents. In our method, we approximate hormonal secretory events by deconvolving GH data via a two-step coordinate descent approach. The first step utilizes a sparse-recovery approach to estimate the timing and amplitude of GH secretory events. The second step estimates physiological parameters. Our method identifies the timing and amplitude of GH pulses and system parameters from experimental and simulated data, with a median R2 of 0.93, among experimental data. Recovering GH pulses and model parameters using this approach may improve the quantification of GH parameters under different physiological and pathological conditions and the design and monitoring of interventions.

Robust Resource Allocation for Lightweight Secure Transmission in Multicarrier NOMA-Assisted Full Duplex IoT Networks

In this article, with the aim to enhance the secure transmission and improve the utilization of spectrum resources in Internet of Things (IoT), a multicarrier nonorthogonal multiple access (MC-NOMA)-assisted full duplex (FD) network is investigated, in which nonorthogonal multiple access (NOMA) is implemented in both uplink and downlink transmissions. The lightweight and low-power physical layer security (PLS) technology is employed to protect the information from eavesdropping. Taking the imperfect channel state information (CSI) into account, we formulate a problem to optimize the beamforming vector, artificial noise (AN), transmit power, and subcarrier assignment policy aiming to maximize the worst case sum secrecy rate under the Quality of Service (QoS) and power consumption constraints. Since the formulated problem is nonconvex and difficult to be solved, we decompose it into two joint optimization subproblems. The first is resource allocation with given subcarrier assignment, which is solved by using the block coordinate descent (BCD) approach. The second is subcarrier assignment solved by the matching theory. Our simulation shows that the proposed scheme is robust against the CSI imperfectness of the eavesdropping and self-interference channels, while providing significant sum secrecy rate improvement compared with the orthogonal multiple access (OMA), half duplex (HD) systems, and other benchmark schemes.

Optimized economic operation of energy storage integration using improved gravitational search algorithm and dual stage optimization

This paper presents optimized energy storage and distribution network model under uncertainty of the power generation problem, based on an improved gravitational search algorithm (IGSA). Throughout the search phase, the IGSA uses a trial-and-error strategy to update the best agent. It also modifies the orbit of the poor agent in the late stages of the search and uses the coordinate descent approach to search for the optimal agent's position. A dual-stage optimization (DSO) model is employed for energy storage and uncertainty of energy generation. The hybrid IGSA-DSO used an adoptive speed inertia coefficient which ensures global and local searchability. The features of the proposed model were validated using IEEE-RTS 24 bus test system with various cases to realize the coordinated planning of the power grid and energy storage. Finally, the simulation shows interesting results of the proposed approach (IGSA-DSO), which reduces the investment cost by 0.248% comparatively.

A non-alternating graph hashing algorithm for large-scale image search

In the era of big data, methods for improving memory and computational efficiency have become crucial for the successful deployment of technology. Hashing is one of the most effective approaches to deal with the computational limitations associated with big data. One natural way to formulate this problem is spectral hashing, which directly incorporates an affinity to learning binary codes. However, owing to the binary constraints, the optimization becomes intractable. To mitigate this challenge, different relaxation approaches have been proposed to reduce the computational load required to obtain binary codes and still attain a good solution. The problem with all existing relaxation methods involves the use of one or more additional auxiliary variables to attain high-quality binary codes while relaxing the problem. The existence of auxiliary variables leads to the coordinate descent approach, which increases the computational complexity. We argue that the introduction of these variables is unnecessary. To this end, we propose a novel relaxed formulation for spectral hashing that adds no additional variables to the problem. Furthermore, instead of solving the problem in the original space where the number of variables is equal to the data points, we solve the problem in a much smaller space and retrieve the binary codes from this solution. This technique reduces both the memory and computational complexity simultaneously. We apply two optimization techniques, namely, the projected gradient and optimization on the manifold, to obtain the solution. Using comprehensive experiments on four public datasets, we show that the proposed efficient spectral hashing (ESH) algorithm achieves a highly competitive retrieval performance compared with the state-of-the-art algorithms at low complexity.

Secrecy Energy Efficiency Maximization in UAV-Enabled Wireless Sensor Networks Without Eavesdropper’s CSI

Unmanned aerial vehicles (UAVs) are anticipated to be a potential data collection solution for wireless sensor networks (WSNs). The main challenges of integrating UAVs in WSNs are security threats and UAV’s onboard energy limitation. To cope with these two challenges, this article examines the secrecy energy efficiency (SEE) maximization problem in UAV-enabled WSN. Specifically, a full-duplex (FD) UAV gathers confidential information from ground sensor nodes (SNs) in the uplink while sending jamming signals to confound a ground eavesdropper (Eve) in the downlink. Considering a passive eavesdropping scenario lacking Eve’s instantaneous channel state information (CSI), the resulting problem is subject to the constraints of connection outage probability (COP), secrecy outage probability (SOP), securely collected bits, and flight trajectory. To tackle the intractable nonconvex problem, we first derive the optimal codeword rate and redundancy rate in closed-form expressions and then develop a low-complexity algorithm using the block coordinate descent (BCD) approach to alternatively optimize the SN scheduling, SN transmit power, UAV transmit power, and UAV trajectory. Simulation results verify the performance gains of the proposed scheme compared with the benchmark schemes. In particular, the proposed scheme achieves nearly the same secrecy rate gains at a lower UAV’s energy consumption cost than the sum secrecy rate maximization (SSRM) baseline. Moreover, it is revealed that trajectory optimization of the proposed scheme plays a crucial role in improving SEE performance compared with the circle trajectory (CT) baseline.

Pliable lasso for the support vector machine

In this article, we study the support vector machine with interaction effects. The pliable lasso penalty, which allows for estimating the main effects of the covariates X and the interaction effects between the covariates and a set modifiers Z is implemented to handle the interaction effect. Interaction variables are included in a hierarchical manner by first considering whether their corresponding main effect variables have been included in the model to avoid over-fitting. The loss function employed is the squared hinge loss, with the pliable lasso penalty and then, the block-wise coordinate descent approach is employed. The results from the simulation and real data show the effectiveness of the pliable lasso in building support vector machine models in situations where interaction effects are involved.

Beyond neighbourhood-preserving transformations for quantization-based unsupervised hashing

• A new type of non-rigid transformation is presented for quantization based hashing. • The proposed non-rigid transformation corrupts the neighborhood structure of data in favor of quantization. • The proposed method reduces dimensionality and quantization error simultaneously. • Unlike other methods, we employ both non-rigid and rigid transformations to compress the data. • An efficient sequential scheme is presented to update optimization variables with low training time. An effective unsupervised hashing algorithm leads to compact binary codes preserving the neighborhood structure of data as much as possible. One of the most established schemes for unsupervised hashing is to reduce the dimensionality of data and then find a rigid (neighborhood-preserving) transformation that reduces the quantization error. Although employing rigid transformations is effective, we may not reduce quantization loss to the ultimate limits. As well, reducing dimensionality and quantization loss in two separate steps seems to be sub-optimal. Motivated by these shortcomings, we propose to employ both rigid and non-rigid transformations to reduce quantization error and dimensionality simultaneously. We relax the orthogonality constraint on the projection in a PCA-formulation and regularize this by a quantization term. We show that both the non-rigid projection matrix and rotation matrix contribute towards minimizing quantization loss but in different ways. A scalable nested coordinate descent approach is proposed to optimize this mixed-integer optimization problem. We evaluate the proposed method on five public benchmark datasets providing almost half a million images. Comparative results indicate that the proposed method mostly outperforms state-of-art linear methods and competes with end-to-end deep solutions.

TU85. POLYGENIC RISK SCORE FOR CANNABIS USE DISORDER PREDICTS ACUTE INTOXICATING EFFECTS AND COGNITIVE DEFICITS INDUCED BY DELTA-9-TETRAHYDROCANNABINOL: A PILOT STUDY

Algorithms for learning overcomplete dictionaries for sparse signal representation are mostly iterative minimization methods that alternate between a sparse coding stage and a dictionary update stage. For most however, the notion of consistency of the learned quantities has not been addressed. Based on the observation that the observed signals can be approximated as a sum of rank one matrices, a new adaptive dictionary learning algorithm is proposed in this paper. It is derived via sequential adaptive penalized rank one matrix approximation where the ℓ1-norm is introduced as a penalty promoting sparsity. The proposed algorithm uses a block coordinate descent approach to consistently estimate the unknowns and has the advantage of having simple closed form solutions for both the sparse coding and dictionary update stages. The consistency properties of both the estimated sparse code and dictionary atom are provided. The performance of the proposed algorithm compared to some state of the art algorithms is illustrated on both simulated data and a real functional magnetic resonance imaging (fMRI) data set from a finger tapping experiment.

100% saturated liquid hydrogen production: Mixed-refrigerant cascaded process with two-stage ortho-to-para hydrogen conversion

To reduce CO2 emissions and address climate change concerns, most futuristic studies investigating 100% renewable energy sources and subsequent power-to-gas/fuel/liquid/X technological developments have been based on hydrogen (H2). The long-term storage and transportation of H2 over long distances restrict its feasibility as an energy vector, mainly due to its low energy density. Liquefaction is a promising approach for overcoming these issues. However, it requires a large amount of energy, and if H2 itself is used to provide this energy, then 25% to 35% of the initial quantity of H2 is consumed. The existing H2 liquefaction plants have specific energy consumption values in the range of 10–12 kWh/kgLH2 and exergy efficiencies in the range of 20%–30% with complicated configurations. Therefore, a thermodynamically efficient and compact design is required to facilitate a roadmap to H2 economy. This paper proposes a simple, energy-efficient, and cost-effective process for H2 liquefaction. Three refrigeration cycles with optimal mixed-refrigerant compositions are used, which makes the proposed process energy-efficient. Additionally, two-stage ortho-to-para conversion makes the process compact. The proposed process is unique in terms of its configuration and mixed-refrigerant combination. The modified coordinate descent approach was adopted to identify the optimal design variables for the proposed H2 liquefaction process. The proposed process consumes an energy of 6.45 kWh/kgLH2, which is 36.5% and 16.1% lower than that consumed by the base design of the proposed process and a published base case, respectively. Additionally, the exergy efficiency of the proposed process is 47.2%. This study will help process engineers achieve a sustainable green economy by improving the competitiveness of H2 storage and transportation over long distances.

A Block Coordinate Descent-Based Projected Gradient Algorithm for Orthogonal Non-Negative Matrix Factorization

This article uses the projected gradient method (PG) for a non-negative matrix factorization problem (NMF), where one or both matrix factors must have orthonormal columns or rows. We penalize the orthonormality constraints and apply the PG method via a block coordinate descent approach. This means that at a certain time one matrix factor is fixed and the other is updated by moving along the steepest descent direction computed from the penalized objective function and projecting onto the space of non-negative matrices. Our method is tested on two sets of synthetic data for various values of penalty parameters. The performance is compared to the well-known multiplicative update (MU) method from Ding (2006), and with a modified global convergent variant of the MU algorithm recently proposed by Mirzal (2014). We provide extensive numerical results coupled with appropriate visualizations, which demonstrate that our method is very competitive and usually outperforms the other two methods.

Identification of Sympathetic Nervous System Activation From Skin Conductance: A Sparse Decomposition Approach With Physiological Priors.

Sweat secretions lead to variations in skin conductance (SC) signal. The relatively fast variation of SC, called the phasic component, reflects sympathetic nervous system activity. The slow variation related to thermoregulation and general arousal is known as the tonic component. It is challenging to decompose the SC signal into its constituents to decipher the encoded neural information related to emotional arousal. We model the phasic component using a second-order differential equation representing the diffusion and evaporation processes of sweating. We include a sparse impulsive neural signal that stimulates the sweat glands for sweat production. We model the tonic component with several cubic B-spline functions. We formulate an optimization problem with physiological priors on system parameters, a sparsity prior on the neural stimuli, and a smoothness prior on the tonic component. Finally, we employ a generalized-cross-validation-based coordinate descent approach to balance among the smoothness of the tonic component, the sparsity of the neural stimuli, and the residual. We illustrate that we can successfully recover the unknowns separating both tonic and phasic components from both experimental and simulated data (with ). Further, we successfully demonstrate our ability to automatically identify the sparsity level for the neural stimuli and the smoothness level for the tonic component. Our generalized-cross-validation-based novel method for SC signal decomposition successfully addresses previous challenges and retrieves a physiologically plausible solution. Accurate decomposition of SC could potentially improve cognitive stress tracking in patients with mental disorders.