Search

With the development of edge devices and cloud computing, the question of how to accomplish machine learning and optimization tasks in a privacy-preserving and secure way has attracted increased attention over the past decade. As a privacy-preserving distributed machine learning method, federated learning (FL) has become popular in the last few years. However, the data privacy issue also occurs when solving optimization problems, which has received little attention so far. This survey paper is concerned with privacy-preserving optimization, with a focus on privacy-preserving data-driven evolutionary optimization. It aims to provide a roadmap from secure privacy-preserving learning to secure privacy-preserving optimization by summarizing security mechanisms and privacy-preserving approaches that can be employed in machine learning and optimization. We provide a formal definition of security and privacy in learning, followed by a comprehensive review of FL schemes and cryptographic privacy-preserving techniques. Then, we present ideas on the emerging area of privacy-preserving optimization, ranging from privacy-preserving distributed optimization to privacy-preserving evolutionary optimization and privacy-preserving Bayesian optimization (BO). We further provide a thorough security analysis of BO and evolutionary optimization methods from the perspective of inferring attacks and active attacks. On the basis of the above, an in-depth discussion is given to analyze what FL and distributed optimization strategies can be used for the design of federated optimization and what additional requirements are needed for achieving these strategies. Finally, we conclude the survey by outlining open questions and remaining challenges in federated data-driven optimization. We hope this survey can provide insights into the relationship between FL and federated optimization and will promote research interest in secure federated optimization.

Abstract Background MCL-1 is a prosurvival B-cell lymphoma 2 family protein that plays a critical role in tumor maintenance and survival and can act as a resistance factor to multiple anticancer therapies. Herein, we describe the generation and characterization of the highly potent and selective MCL-1 inhibitor ABBV-467 and present findings from a first-in-human trial that included patients with relapsed/refractory multiple myeloma (NCT04178902). Methods Binding of ABBV-467 to human MCL-1 was assessed in multiple cell lines. The ability of ABBV-467 to induce tumor growth inhibition was investigated in xenograft models of human multiple myeloma and acute myelogenous leukemia. The first-in-human study was a multicenter, open-label, dose-escalation study assessing safety, pharmacokinetics, and efficacy of ABBV-467 monotherapy. Results Here we show that administration of ABBV-467 to MCL-1-dependent tumor cell lines triggers rapid and mechanism-based apoptosis. In vivo, intermittent dosing of ABBV-467 as monotherapy or in combination with venetoclax inhibits the growth of xenografts from human hematologic cancers. Results from a clinical trial evaluating ABBV-467 in patients with multiple myeloma based on these preclinical data indicate that treatment with ABBV-467 can result in disease control (seen in 1 patient), but may also cause increases in cardiac troponin levels in the plasma in some patients (seen in 4 of 8 patients), without other corresponding cardiac findings. Conclusions The selectivity of ABBV-467 suggests that treatment-induced troponin release is a consequence of MCL-1 inhibition and therefore may represent a class effect of MCL-1 inhibitors in human patients.

Abstract In many real-world applications of interest, several related optimization tasks can be encountered, where each task is associated with a specific context or personalized information. Moreover, the amount of available data for each task may be highly limited due to the expensive cost involved. Although Bayesian optimization (BO) has emerged as a promising paradigm for handling black-box optimization problems, addressing such a sequence of optimization tasks can be intractable due to the cold start issues in BO. The key challenge is to speed up the optimization by leveraging the transferable information, while taking the personalization into consideration. In this paper, optimization problems with personalized variables are formally defined at first. Subsequently, a personalized evolutionary Bayesian algorithm is proposed to consider the personalized information and the measurement noise. Specifically, a contextual Gaussian process is used to jointly learn a surrogate model in different contexts with regard to the varying personalized parameter, and an evolutionary algorithm is tailored for optimizing an acquisition function for handling the presence of personalized information. Finally, we demonstrate the effectiveness of the proposed algorithm by testing it on widely used single- and multi-objective benchmark problems with personalized variables.

In high-energy astrophysical processes involving compact objects, such as core-collapse supernovae or binary neutron star mergers, neutrinos play an important role in the synthesis of nuclides. Neutrinos in these environments can experience collective flavor oscillations driven by neutrino–neutrino interactions, including coherent forward scattering and incoherent (collisional) effects. Recently, there has been interest in exploring potential novel behaviors in collective oscillations of neutrinos by going beyond the one-particle effective or “mean-field” treatments. Here, we seek to explore implications of collective neutrino oscillations, in the mean-field treatment and beyond, for the nucleosynthesis yields in supernova environments with different astrophysical conditions and neutrino inputs. We find that collective oscillations can impact the operation of the ν p -process and r -process nucleosynthesis in supernovae. The potential impact is particularly strong in high-entropy, proton-rich conditions, where we find that neutrino interactions can nudge an initial ν p -process neutron-rich, resulting in a unique combination of proton-rich low-mass nuclei as well as neutron-rich high-mass nuclei. We describe this neutrino-induced neutron-capture process as the “ ν i -process.” In addition, nontrivial quantum correlations among neutrinos, if present significantly, could lead to different nuclide yields compared to the corresponding mean-field oscillation treatments, by virtue of modifying the evolution of the relevant one-body neutrino observables.

Binary neutron star mergers produce high-energy emissions from several physically different sources, including a gamma-ray burst (GRB) and its afterglow, a kilonova (KN), and, at late times, a remnant many parsecs in size. Ionizing radiation from these sources can be dangerous for life on Earth-like planets when located too close. Work to date has explored the substantial danger posed by the GRB to on-axis observers; here we focus instead on the potential threats posed to nearby off-axis observers. Our analysis is based largely on observations of the GW170817/GRB 170817A multi-messenger event, as well as theoretical predictions. For baseline KN parameters, we find that the X-ray emission from the afterglow may be lethal out to ∼1 pc and the off-axis gamma-ray emission may threaten a range out to ∼4 pc, whereas the greatest threat comes years after the explosion, from the cosmic rays accelerated by the KN blast, which can be lethal out to distances up to ∼11 pc. The distances quoted here are typical, but the values have significant uncertainties and depend on the viewing angle, ejected mass, and explosion energy in ways we quantify. Assessing the overall threat to Earth-like planets, KNe have a similar kill distance to supernovae, but are far less common. However, our results rely on the scant available KN data, and multi-messenger observations will clarify the danger posed by such events.

Tinnitus is perception of sound in the absence of an apparent external acoustic stimulus. The condition is prevalent in adults, especially the elderly (≥65 years), and may be associated with cognitive function decline and significantly impacts on the quality of life, heralding difficulties in managing this challenging disorder. Interventions for tinnitus have been varied. However, drugs have not yet been approved for the treatment of tinnitus and there is no pharmacotherapy recommended by existing guidelines. Still, herbal medicines are used for the treatment of tinnitus in many countries, especially Gingko (G.) biloba. In the current updated literature review, we evaluated the efficacy of herbal medicines in the treatment of tinnitus by reviewing the evidence of relevant randomized controlled trials. The authors also highlight some of the issues in clinical trials of herbal medicines given that currently available evidence on herbal medicines for tinnitus is overall of insufficient quality and the conclusions from existing trials are conflicting. Nevertheless, there is a clear and urgent need for safe and effective pharmacotherapy of tinnitus.

Administration of cisplatin, a common chemotherapeutic drug, has an inevitable side effect of sensorineural hearing loss. The main etiologies are stria vascularis injury, spiral ganglion degeneration, and hair cell death. Over several decades, the research scope of cisplatin-induced ototoxicity has expanded with the discovery of the molecular mechanism mediating inner ear cell death, highlighting the roles of reactive oxygen species and transport channels for cisplatin uptake into inner ear cells. Upon entering hair cells, cisplatin disrupts organelle metabolism, induces oxidative stress, and targets DNA to cause intracellular damage. Recent studies have also reported the role of inflammation in cisplatin-induced ototoxicity. In this article, we preform a narrative review of the latest reported molecular mechanisms of cisplatin-induced ototoxicity, from extracellular to intracellular. We build up a signaling network starting with cisplatin entering into the inner ear through the blood labyrinth barrier, disrupting cochlear endolymph homeostasis, and activating inflammatory responses of the outer hair cells. After entering the hair cells, cisplatin causes hair cell death via DNA damage, redox system imbalance, and mitochondrial and endoplasmic reticulum dysfunction, culminating in programmed cell death including apoptosis, necroptosis, autophagic death, pyroptosis, and ferroptosis. Based on the mentioned mechanisms, prominent therapeutic targets, such as channel-blocking drugs of cisplatin transporter, construction of cisplatin structural analogues, anti-inflammatory drugs, antioxidants, cell death inhibitors, and others, were collated. Considering the recent research efforts, we have analyzed the feasibility of the aforementioned therapeutic strategies and proposed our otoprotective approaches to overcome cisplatin-induced ototoxicity.

Long-lived massive magnetars are expected to be remnants of some binary neutron star (BNS) mergers. In this paper, we argue that the magnetic powered flaring activities of these merged magnetars would occur dominantly in their early millisecond-period-spin phase, which is in the timescale of days. Such flares endure significant absorption by the ejecta from the BNS collision, and their detectable energy range is from 0.1 to 10 MeV, in a time lag of approximately days after the merger events indicated by the gravitational wave chirps. We estimate the rate of such flares in different energy ranges, and find that there could have been 0.1–10 cases detected by Fermi/GBM. A careful search for ∼10 ms spin-period modulation in weak short gamma-ray bursts (GRBs) may identify them from the archival data. The next-generation megaelectronvolt detectors could detect them at a mildly higher rate. The recent report on the Quasi-Period-Oscillation found in two BASTE GRBs should not be considered as cases of such flares, for they were detected in a lower energy range and with a much shorter period spin modulation.

^244 Pu has recently been discovered in deep-sea deposits spanning the past 10 Myr, a period that includes two ^60 Fe pulses from nearby supernovae. ^244 Pu is among the heaviest r -process products, and we consider whether it was created in supernovae, which is disfavored by nucleosynthesis simulations, or in an earlier kilonova event that seeded the nearby interstellar medium with ^244 Pu that was subsequently swept up by the supernova debris. We discuss how these possibilities can be probed by measuring ^244 Pu and other r -process radioisotopes such as ^129 I and ^182 Hf, both in lunar regolith samples returned to Earth by missions such as Chang’e and Artemis, and in deep-sea deposits.

Particle filters, also known as sequential Monte Carlo (SMC) methods, constitute a class of importance sampling and resampling techniques designed to use simulations to perform on-line filtering. Recently, particle filters have been extended for optimization by utilizing the ability to track a sequence of distributions. In this work, we incorporate transfer learning capabilities into the optimizer by using particle filters. To achieve this, we propose a novel particle-filter-based multi-objective optimization algorithm (PF-MOA) by transferring knowledge acquired from the search experience. The key insight adopted here is that, if we can construct a sequence of target distributions that can balance the multiple objectives and make the degree of the balance controllable, we can approximate the Pareto optimal solutions by simulating each target distribution via particle filters. As the importance weight updating step takes the previous target distribution as the proposal distribution and takes the current target distribution as the target distribution, the knowledge acquired from the previous run can be utilized in the current run by carefully designing the set of target distributions. The experimental results on the DTLZ and WFG test suites show that the proposed PF-MOA achieves competitive performance compared with state-of-the-art multi-objective evolutionary algorithms on most test instances.

Bayesian optimization has emerged at the forefront of expensive black-box optimization due to its data efficiency. Recent years have witnessed a proliferation of studies on the development of new Bayesian optimization algorithms and their applications. Hence, this article attempts to provide a comprehensive and updated survey of recent advances in Bayesian optimization that are mainly based on Gaussian processes and identify challenging open problems. We categorize the existing work on Bayesian optimization into nine main groups according to the motivations and focus of the proposed algorithms. For each category, we present the main advances with respect to the construction of surrogate models and adaptation of the acquisition functions. Finally, we discuss the open questions and suggest promising future research directions, in particular with regard to heterogeneity, privacy preservation, and fairness in distributed and federated optimization systems. [ABSTRACT FROM AUTHOR]

In this paper, we propose a deep neural network-based numerical method for solving contact problems. Focusing on a static frictionless unilateral contact problem, we derive its weak formulation and prove that the solution of the weak formulation is also the minimizer of the corresponding energy functional. By converting the original contact problem into a minimization problem, a deep neural network is adopted to approximate the solution and solve the minimization problem. Numerical results demonstrate the effectiveness and accuracy of our method. [ABSTRACT FROM AUTHOR]

A solution to layout problems of artificial satellite modules was studied and developed. The problem, which is classified as NP-complete, involves the search for an optimal method to install on a base board a number of two-dimensional graph elements. The installation, however, requires no overlapping of graph elements, the size of the graph elements not to exceed the region and minimal distances between the base board and centroid of all pieces. A new algorithm of global optimization was also presented to respond to the problem.

In this paper, we consider a mathematical model which describes the quasistatic frictionless contact between a viscoplastic body and a foundation. The contact is modeled with normal compliance and unilateral constraint. We present the variational-hemivariational formulation of the model and prove its unique solvability. Then we introduce a fully discrete scheme to solve the problem and derive an error estimate. Under appropriate regularity assumptions of the exact solution, we obtain the optimal order error estimate. Finally, numerical results are reported to show the performance of the numerical method. [ABSTRACT FROM AUTHOR]

A Milky Way Type Ia supernova (SNIa) could be unidentified or even initially unnoticed, being dim in radio, X-rays, and neutrinos, and suffering large optical/IR extinction in the Galactic plane. But SNIa emit nuclear gamma-ray lines from 56Ni → 56Co → 56Fe radioactive decays. These lines fall within the Fermi /GBM energy range, and the 56Ni 158 keV line is detectable by Swift /BAT. Both instruments frequently monitor the Galactic plane, which is transparent to gamma rays. Thus GBM and BAT are ideal Galactic SNIa early warning systems. We simulate SNIa MeV light curves and spectra to show that GBM and BAT could confirm a Galactic SNIa explosion, followed by Swift localization and observation in X-rays and UVOIR band. The time of detection depends sensitively on the 56Ni distribution, and can be as early as a few days if |${\gtrsim } 10{{\ \rm per\ cent}}$| of the 56Ni is present in the surface as suggested by SN2014J gamma data. [ABSTRACT FROM AUTHOR]

Several starburst galaxies have been observed in the GeV and TeV bands. In these dense environments, gamma-ray emission should be dominated by cosmic ray (CR) interactions with the interstellar medium (pcrpism → π0 → γγ). Indeed, starbursts may act as proton 'calorimeters' where a substantial fraction of CR energy input is emitted in gamma-rays. Here, we build a one-zone, 'thick-target' model implementing calorimetry and placing a firm upper bound on gamma-ray emission from CR interactions. The model assumes that CRs are accelerated by supernovae (SNe), and all suffer nuclear interactions rather than escape. Our model has only two free parameters: the CR proton acceleration energy per SN ϵcr, and the proton injection spectral index s. We calculate the pionic gamma-ray emission from 10 MeV to 10 TeV, and derive thick-target parameters for six galaxies with Fermi, H.E.S.S., and/or VERITAS data. Our model provides good fits for the M82 and NGC 253, and yields ϵcr and s values suggesting that SN CR acceleration is similar in starbursts and in our Galaxy. We find that these starbursts are indeed nearly if not fully proton calorimeters. For NGC 4945 and NGC 1068, the models are consistent with calorimetry but are less well-constrained due to the lack of TeV data. However, the Circinus galaxy and the ultra-luminous infrared galaxy Arp 220 exceed our pionic upper-limit; possible explanations are discussed. [ABSTRACT FROM AUTHOR]

Rapid advancement of sensor technologies and computing power has led to wide availability of massive population-based shape data. In this paper, we present a Taylor expansion-based method for computing structural performance variation over its shape population. The proposed method consists of four steps: (1) learning the shape parameters and their probabilistic distributions through the statistical shape modeling (SSM), (2) deriving analytical sensitivity of structural performance over shape parameter, (3) approximating the explicit function relationship between the finite element (FE) solution and the shape parameters through Taylor expansion, and (4) computing the performance variation by the explicit function relationship. To overcome the potential inaccuracy of Taylor expansion for highly nonlinear problems, a multipoint Taylor expansion technique is proposed, where the parameter space is partitioned into different regions and multiple Taylor expansions are locally conducted. It works especially well when combined with the dimensional reduction of the principal component analysis (PCA) in the statistical shape modeling. Numerical studies illustrate the accuracy and efficiency of this method. [ABSTRACT FROM AUTHOR]

In this paper, we are concerned with an optimal control problem of a quasistatic frictional contact model with history-dependent operators. The contact boundary of the model is divided into two parts where different contact conditions are specified. For the contact problem, we first derive its weak formulation and prove the existence and uniqueness of the solution to the weak formulation. Then we give a priori estimate of the unique solution and prove a continuous dependence result for the solution map. Finally, an optimal control problem that contains boundary and initial condition controls is proposed, and the existence of optimal solutions to the control problem is established. [ABSTRACT FROM AUTHOR]

The molecular chaperone HSP90 has been shown to facilitate cancer cell survival by stabilizing key proteins responsible for a malignant phenotype. We report here the results of parallel fragment-based drug design approaches in the design of novel HSP90 inhibitors. Initial aminopyrimidine leads were elaborated using high-throughput organic synthesis to yield nanomolar inhibitors of the enzyme. Second site leads were also identified which bound to HSP90 in two distinct conformations, an ‘open’ and ‘closed’ form. Intriguingly, linked fragment approaches targeting both of these conformations were successful in producing novel, micromolar inhibitors. Overall, this study shows that, with only a few fragment hits, multiple lead series can be generated for HSP90 due to the inherent flexibility of the active site. Thus, ample opportunities exist to use these lead series in the development of clinically useful HSP90 inhibitors for the treatment of cancers. [ABSTRACT FROM AUTHOR]