Search

In this article, we consider the problem of unconstrained time-varying convex optimization, where the cost function changes with time. We provide an in-depth technical analysis of the problem and argue why freezing the cost at each time step and taking finite steps toward the minimizer is not the best tracking solution for this problem. We propose a set of algorithms that by taking into account the temporal variation of the cost aim to reduce the tracking error of the time-varying minimizer of the problem. The main contribution of our work is that our proposed algorithms only require the first-order derivatives of the cost function with respect to the decision variable. This approach significantly reduces computational cost compared to the existing algorithms, which use the inverse of the Hessian of the cost. Specifically, the proposed algorithms reduce the computational cost from $O(n^3)$ to $O(n)$ per timestep, where $n$ is the size of the decision variable. Avoiding the inverse of the Hessian also makes our algorithms applicable to non-convex optimization problems. We refer to these algorithms as $O(n)$-algorithms. These $O(n)$-algorithms are designed to solve the problem for different scenarios based on the available temporal information about the cost. We illustrate our results through various examples, including the solution of a model predictive control problem framed as a convex optimization problem with a streaming time-varying cost function., Comment: arXiv admin note: text overlap with arXiv:2310.07925

Federated learning (FL) has gained considerable popularity for distributed machine learning due to its ability to preserve the privacy of participating agents by eliminating the need for data aggregation. Nevertheless, communication costs between agents and the central server in FL are substantial in large-scale problems and remain a limiting factor for this algorithm. This paper introduces an innovative algorithm, called \emph{FedScalar}, within the federated learning framework aimed at improving communication efficiency. Unlike traditional FL methods that require agents to send high-dimensional vectors to the server, \emph{FedScalar} enables agents to communicate updates using a single scalar. Each agent encodes its updated model parameters into a scalar through the inner product between its local update difference and a random vector, which is then transmitted to the server. The server decodes this information by projecting the averaged scalar values onto the random vector. Our method thereby significantly reduces communication overhead. Technically, we demonstrate that the proposed algorithm achieves a convergence rate of $O(1/\sqrt{K})$ to a stationary point for smooth, non-convex loss functions. Additionally, our analysis shows that altering the underlying distribution of the random vector generated by the server can reduce the variance during the aggregation step of the algorithm. Finally, we validate the performance and communication efficiency of our algorithm with numerical simulations.

This paper aims to enhance the use of the Frank-Wolfe (FW) algorithm for training deep neural networks. Similar to any gradient-based optimization algorithm, FW suffers from high computational and memory costs when computing gradients for DNNs. This paper introduces the application of the recently proposed projected forward gradient (Projected-FG) method to the FW framework, offering reduced computational cost similar to backpropagation and low memory utilization akin to forward propagation. Our results show that trivial application of the Projected-FG introduces non-vanishing convergence error due to the stochastic noise that the Projected-FG method introduces in the process. This noise results in an non-vanishing variance in the Projected-FG estimated gradient. To address this, we propose a variance reduction approach by aggregating historical Projected-FG directions. We demonstrate rigorously that this approach ensures convergence to the optimal solution for convex functions and to a stationary point for non-convex functions. These convergence properties are validated through a numerical example, showcasing the approach's effectiveness and efficiency.

This paper proposes a set of novel optimization algorithms for solving a class of convex optimization problems with time-varying streaming cost function. We develop an approach to track the optimal solution with a bounded error. Unlike the existing results, our algorithm is executed only by using the first-order derivatives of the cost function which makes it computationally efficient for optimization with time-varying cost function. We compare our algorithms to the gradient descent algorithm and show why gradient descent is not an effective solution for optimization problems with time-varying cost. Several examples including solving a model predictive control problem cast as a convex optimization problem with a streaming time-varying cost function demonstrate our results.

We prove that every profinite $n$-ary group $(G, f)=\Gf$ has a unique Haar measure $m_p$ and further for every measurable subset $A\subseteq G$, we have $$ m_p(A)=m(A)=(n-1)m^{\ast}(A) $$ where $m$ and $m^{\ast}$ are the normalized Haar measures of the profinite groups $(G, \bullet)$ and the Post cover $G^{\ast}$, respectively.

This paper proposes an algorithm for Federated Learning (FL) with a two-layer structure that achieves both variance reduction and a faster convergence rate to an optimal solution in the setting where each agent has an arbitrary probability of selection in each iteration. In distributed machine learning, when privacy matters, FL is a functional tool. Placing FL in an environment where it has some irregular connections of agents (devices), reaching a trained model in both an economical and quick way can be a demanding job. The first layer of our algorithm corresponds to the model parameter propagation across agents done by the server. In the second layer, each agent does its local update with a stochastic and variance-reduced technique called Stochastic Variance Reduced Gradient (SVRG). We leverage the concept of variance reduction from stochastic optimization when the agents want to do their local update step to reduce the variance caused by stochastic gradient descent (SGD). We provide a convergence bound for our algorithm which improves the rate from $O(\frac{1}{\sqrt{K}})$ to $O(\frac{1}{K})$ by using a constant step-size. We demonstrate the performance of our algorithm using numerical examples.

We study the structure of profinite polyadic groups and we prove that a polyadic topological group $(G, f)$ is profinite, if and only if, it is compact, Hausdorff, totally disconnected. More generally, for a pseudo-variety (or a formation) of finite groups $\mathfrak{X}$, we define the class of $\mathfrak{X}$-polyadic groups, and we show that a polyadic group $(G, f)$ is pro-$\mathfrak{X}$, if and only if, it is compact, Hausdorff, totally disconnected and for every open congruence $R$, the quotient $(G/R, f_R)$ is $\mathfrak{X}$-polyadic., Comment: 11 pages

Given a large graph, a graph sample determines a subgraph with similar characteristics for certain metrics of the original graph. The samples are much smaller thereby accelerating and simplifying the analysis and visualization of large graphs. We focus on the implementation of distributed graph sampling for Big Data frameworks and in-memory dataflow systems such as Apache Spark or Apache Flink. We evaluate the scalability of the new implementations and analyze to what degree the sampling approaches preserve certain graph metrics compared to the original graph. The latter analysis also uses comparative graph visualizations. The presented methods will be open source and be integrated into Gradoop, a system for distributed graph analytics.

In this paper, we investigate the critical points of the $ P-V $ diagram and the phase transitions of the Horndenski black holes. In fact, the usual Horndeski black holes do not have $ P-V $ critical points, hence do not show any phase transitions. However, we successfully modify the Horndeski black hole solution to obtain such a phase transition behavior. This modified black holes solution is satisfied by the equation of state of liquid-gas phase transition. Also, we study the thermodynamics of our modified Horndeski black hole by applying a new method based on the equation of state that originated from the slope of temperature versus entropy. This new prescription provides us a simple powerful way to study the critical behavior and the phase transition of the black holes and concludes the novel results. The analytical interpretation of possible phase transition points leads us to set some nonphysical range on the horizon radius for the black hole., Comment: 18 pages, 8 figures

In this paper, the thermodynamical properties and the phase transitions of the charged accelerating anti-de Sitter (AdS) black holes are investigated in the framework of the $f(R)$ gravity. By studying the conditions for the phase transitions, it has been shown that the $P-V$ criticality and the van der Waals like phase transitions can be achieved for $ T \approx T_{c} $. The Joule-Thomson expansion effects are also examined for the charged accelerating AdS black holes of the $f(R)$ gravity. Here, we derive the inversion temperatures as well as the inversion curves. Then, we determine the position of the reverse point for different values of mass $M$ and parameter $b$ for the corresponding black hole. At this point, the Joule-Thompson coefficient is zero. So, in such case, we can say that such point is very important for the finding of cooling - heating regions. Finally, we calculate the ratio of minimum inversion temperature and critical temperature for such black hole., Comment: 19 pages, 7 figures

In this paper, we study the notion of $\phi$-biflatness for some Banach algebras, where $\phi$ is a non-zero multiplicative linear functional. We show that the Segal algebra $S(G)$ is left $\phi$-biflat if and only if $G$ is amenable. Also, we characterize left $\phi$-biflatness of semigroup algebra $\ell^{1}(S)$ in the term of biflatness, where $S$ is a Clifford semigroup.

In this paper, we consider the first order correction of the entropy and temperature in a charged black hole with a scalar field. Here, we apply such correction for the different cases of black holes. These corrections are due to the thermal fluctuations of statistical physics. Also, we take advantage of such corrections and study the $ P - V $ critically and phase transition. Also, we investigate the effect of correction on the critical point and stability of the system. We obtain modified thermodynamics quantities and find effects of thermal fluctuations in the stability of the black hole. We find that stability of black hole is depend on the thermal fluctuations. Finally, we compare the results of the corrected and uncorrected by thermodynamical quantities., Comment: 23 pages, 13 figures

In this paper, we study the thermodynamic properties of Kiselev black hole and its holographic dual. We obtain the thermodynamic product formula for the Kiselev black hole. We consider Kiselev Black hole surrounded by both dust and radiation. We find that the area (or entropy) product formula for both cases is mass-independent as well as the case in the Einstein gravity, which interpreted as universal quantity. Moreover, we calculate the black hole entropy bound for both the inner and outer horizons. Furthermore, we show that the central charges of the left and right moving sectors are not the same via universal thermodynamic relations. Such universal relations lead us to calculate the central charge of conformal field theory (CFT). Finally, we see that the left and right of central charge of CFT are same., Comment: 17 pages, References added in the new version

In this paper, we consider dyonic charged AdS black hole which is holographic dual of a van der Waals fluid. We use logarithmic corrected entropy and study thermodynamics of the black hole and show that holographic picture is still valid. Critical behaviors and stability also discussed. Logarithmic corrections arises due to thermal fluctuations which are important when size of black hole be small. So, thermal fluctuations interpreted as quantum effect. It means that we can see quantum effect of a black hole which is a gravitational system., Comment: 11 pages, 5 figures. References added. Revised according to PRD report

In this paper, we consider an inflationary model of $f(R)$ gravity with polynomial form plus logarithmic term. We calculate some cosmological parameters and compare our results with the Plank 2015 data. We find that presence of both logarithmic and polynomial corrections are necessary to yield slow-roll condition. Also, we study critical points and stability of the model to find that it is a viable model., Comment: 15 pages, 9 figures

Let P and Q be convex polyhedra in E3 with face lattices F(P) and F(Q) and symmetry groups G(P) and G(Q), respectively. Then, P and Q are called face equivalent if there is a lattice isomorphism between F(P) and F(Q); P and Q are called symmetry equivalent if the action of G(P) on F(P) is equivalent to the action of G(Q) on F(Q). It is well known that the set [P] of all polyhedra which are face equivalent to P has the structure of a manifold of dimension {e-1}, up to similarities, where e=e(P) is the number of edges of P. This is a consequence of the Steinitz's classical Theorem. We give a new proof of this fact. The symmetry type of P denoted by is the set of all polyhedron Q symmetry equivalent to P. We show that for polyhedra with symmetry group G(P) a reflection group the dimension of this manifold is {O-1} where O is the number of edge orbits of P under and the action of G(P) on F(P).