Differentially private stochastic gradient descent algorithm for multiparty classification

We consider the problem of developing privacy-preserving machine learning algorithms in a dis-tributed multiparty setting. Here different parties own different parts of a data set, and the goal is to learn a classifier from the entire data set with-out any party revealing any information about the individual data points it owns. Pathak et al [7]recently proposed a solution to this problem in which each party learns a local classifier from its own data, and a third party then aggregates these classifiers in a privacy-preserving manner using a cryptographic scheme. The generaliza-tion performance of their algorithm is sensitive to the number of parties and the relative frac-tions of data owned by the different parties. In this paper, we describe a new differentially pri-vate algorithm for the multiparty setting that uses a stochastic gradient descent based procedure to directly optimize the overall multiparty ob-jective rather than combining classifiers learned from optimizing local objectives. The algorithm achieves a slightly weaker form of differential privacy than that of [7], but provides improved generalization guarantees that do not depend on the number of parties or the relative sizes of the individual data sets. Experimental results corrob-orate our theoretical findings.

Fuel-optimal G-MPSP Guidance for Powered Descent Phase of Soft Lunar Landing

A fuel optimal nonlinear sub-optimal guidance scheme is presented in this paper for soft landing of a lunar craft during the powered descent phase. The recently developed Generalized Model Predictive Static Programming (G-MPSP) is used to compute the required magnitude and angle of the thrust vector. Both terminal position and velocity vector are imposed as hard constraints, which ensures high position accuracy and facilitates initiation of vertical descent at the end of the powered descent phase. A key feature of the G-MPSP algorithm is that it converts the nonlinear dynamic programming problem into a low-dimensional static optimization problem (of the same dimension as the output vector). The control history update is done in closed form after computing a time-varying weighting matrix through a backward integration process. This feature makes the algorithm computationally efficient, which makes it suitable for on-board applications. The effectiveness of the proposed guidance algorithm is demonstrated through promising simulation results.

A Fuzzy Approximation Scheme for Sequential Learning in Pattern Recognition

An adaptive learning scheme, based on a fuzzy approximation to the gradient descent method for training a pattern classifier using unlabeled samples, is described. The objective function defined for the fuzzy ISODATA clustering procedure is used as the loss function for computing the gradient. Learning is based on simultaneous fuzzy decisionmaking and estimation. It uses conditional fuzzy measures on unlabeled samples. An exponential membership function is assumed for each class, and the parameters constituting these membership functions are estimated, using the gradient, in a recursive fashion. The induced possibility of occurrence of each class is useful for estimation and is computed using 1) the membership of the new sample in that class and 2) the previously computed average possibility of occurrence of the same class. An inductive entropy measure is defined in terms of induced possibility distribution to measure the extent of learning. The method is illustrated with relevant examples.

Charged-soliton excitations in statistical mechanics

A method is developed for demonstrating how solitons with some internal periodic motion may emerge as elementary excitations in the statistical mechanics of field systems. The procedure is demonstrated in the context of complex scalar fields which can, for appropriate choices of the Lagrangian, yield charge-carrying solitons with such internal motion. The derivation uses the techniques of the steepest-descent method for functional integrals. It is shown that, despite the constraint of some fixed total charge, a gaslike excitation of such charged solitons does emerge.

Effects of environment on surface-wave characteristics of a dielectric-coated conductor .4.

The problem of excitation of 11zultilayercd-graded-dielectric-coatedc onductor by a magnetic ring source is fornzulated in the ,form of a contour integrul which is rolved by using the method of steepest descent. Numerical evaluation of launching efiiency shows that high value of about 90 percent can be attained by choosing proper dimensions of the launcher with respect to the dimension of the surface wave line.

Natural actor-critic algorithms

We present four new reinforcement learning algorithms based on actor-critic, natural-gradient and functi approximation ideas,and we provide their convergence proofs. Actor-critic reinforcement learning methods are online approximations to policy iteration in which the value-function parameters are estimated using temporal difference learning and the policy parameters are updated by stochastic gradient descent. Methods based on policy gradients in this way are of special interest because of their compatibility with function-approximation methods, which are needed to handle large or infinite state spaces. The use of temporal difference learning in this way is of special interest because in many applications it dramatically reduces the variance of the gradient estimates. The use of the natural gradient is of interest because it can produce better conditioned parameterizations and has been shown to further reduce variance in some cases. Our results extend prior two-timescale convergence results for actor-critic methods by Konda and Tsitsiklis by using temporal difference learning in the actor and by incorporating natural gradients. Our results extend prior empirical studies of natural actor-critic methods by Peters, Vijayakumar and Schaal by providing the first convergence proofs and the first fully incremental algorithms.

A Direct Approach to the Problem of Diffraction by a Half-Plane under Mixed Boundary Conditions

A general direct technique of solving a mixed boundary value problem in the theory of diffraction by a semi-infinite plane is presented. Taking account of the correct edge-conditions, the unique solution of the problem is derived, by means of Jones' method in the theory of Wiener-Hopf technique, in the case of incident plane wave. The solution of the half-plane problem is found out in exact form. (The far-field is derived by the method of steepest descent.) It is observed that it is not the Wiener-Hopf technique which really needs any modification but a new technique is certainly required to handle the peculiar type of coupled integral equations which the Wiener-Hopf technique leads to. Eine allgemeine direkte Technik zur Lösung eines gemischten Randwertproblems in der Theorie der Beugung an einer halbunendlichen Ebene wird vorgestellt. Unter Berücksichtigung der korrekten Eckbedingungen wird mit der Methode von Jones aus der Theorie der Wiener-Hopf-Technik die eindeutige Lösung für den Fall der einfallenden ebenen Welle hergeleitet. Die Lösung des Halbebenenproblems wird in exakter Form angegeben. (Das Fernfeld wurde mit der Methode des steilsten Abstiegs bestimmt.) Es wurde bemerkt, daß es nicht die Wiener-Hopf-Technik ist, die wirklich irgend welcher Modifikationen bedurfte. Gewiß aber wird eine neue Technik zur Behandlung des besonderen Typs gekoppelter Integralgleichungen benötigt, auf die die Wiener-Hopf-Technik führt.

Numerical and experimental study of the thermal behaviour of coining and upsetting processes

A numerical simulation technique has been employed to study the thermal behavior of hot-forging type forming processes. Experiments on the coining and upsetting of an aluminum billet were conducted to validate the numerical predictions. Typical forming conditions for both the coining and upsetting processes were then studied in detail. an electrical analogy scheme was used to determine the thermal contact resistance. This scheme can conviniently provide the interface characteristics for typical processing conditions, which normally involve high pressures and temperatures. A single forging cycle was first considered, and then a batch of twenty-five forgings was studied. Each forging cycle includes the billet mounting, ascent, loading, dwelling, unloading, descent, and billet removal stages. The temperature distribution in the first forging to be formed is found to be significantly different from that at the end of the batch. In industry, forging is essentially a batch operation. The influence of forming speed and reduction on thermal characteristics was investigated also. The variations that can occur in the process design by considering differences in temperature characteristics are discussed also.

Variable Sparsity Kernel Learning

This paper(1) presents novel algorithms and applications for a particular class of mixed-norm regularization based Multiple Kernel Learning (MKL) formulations. The formulations assume that the given kernels are grouped and employ l(1) norm regularization for promoting sparsity within RKHS norms of each group and l(s), s >= 2 norm regularization for promoting non-sparse combinations across groups. Various sparsity levels in combining the kernels can be achieved by varying the grouping of kernels-hence we name the formulations as Variable Sparsity Kernel Learning (VSKL) formulations. While previous attempts have a non-convex formulation, here we present a convex formulation which admits efficient Mirror-Descent (MD) based solving techniques. The proposed MD based algorithm optimizes over product of simplices and has a computational complexity of O (m(2)n(tot) log n(max)/epsilon(2)) where m is no. training data points, n(max), n(tot) are the maximum no. kernels in any group, total no. kernels respectively and epsilon is the error in approximating the objective. A detailed proof of convergence of the algorithm is also presented. Experimental results show that the VSKL formulations are well-suited for multi-modal learning tasks like object categorization. Results also show that the MD based algorithm outperforms state-of-the-art MKL solvers in terms of computational efficiency.

On generating optimal signal probabilities for random tests: A genetic approach

Genetic Algorithms are robust search and optimization techniques. A Genetic Algorithm based approach for determining the optimal input distributions for generating random test vectors is proposed in the paper. A cost function based on the COP testability measure for determining the efficacy of the input distributions is discussed, A brief overview of Genetic Algorithms (GAs) and the specific details of our implementation are described. Experimental results based on ISCAS-85 benchmark circuits are presented. The performance pf our GA-based approach is compared with previous results. While the GA generates more efficient input distributions than the previous methods which are based on gradient descent search, the overheads of the GA in computing the input distributions are larger. To account for the relatively quick convergence of the gradient descent methods, we analyze the landscape of the COP-based cost function. We prove that the cost function is unimodal in the search space. This feature makes the cost function amenable to optimization by gradient-descent techniques as compared to random search methods such as Genetic Algorithms.

Efficient algorithms for learning kernels from multiple similarity matrices with general convex loss functions

In this paper we consider the problem of learning an n × n kernel matrix from m(1) similarity matrices under general convex loss. Past research have extensively studied the m = 1 case and have derived several algorithms which require sophisticated techniques like ACCP, SOCP, etc. The existing algorithms do not apply if one uses arbitrary losses and often can not handle m > 1 case. We present several provably convergent iterative algorithms, where each iteration requires either an SVM or a Multiple Kernel Learning (MKL) solver for m > 1 case. One of the major contributions of the paper is to extend the well knownMirror Descent(MD) framework to handle Cartesian product of psd matrices. This novel extension leads to an algorithm, called EMKL, which solves the problem in O(m2 log n 2) iterations; in each iteration one solves an MKL involving m kernels and m eigen-decomposition of n × n matrices. By suitably defining a restriction on the objective function, a faster version of EMKL is proposed, called REKL,which avoids the eigen-decomposition. An alternative to both EMKL and REKL is also suggested which requires only an SVMsolver. Experimental results on real world protein data set involving several similarity matrices illustrate the efficacy of the proposed algorithms.

Two timescale analysis of the Alopex algorithm for optimization

Alopex is a correlation-based gradient-free optimization technique useful in many learning problems. However, there are no analytical results on the asymptotic behavior of this algorithm. This article presents a new version of Alopex that can be analyzed using techniques of two timescale stochastic approximation method. It is shown that the algorithm asymptotically behaves like a gradient-descent method, though it does not need (or estimate) any gradient information. It is also shown, through simulations, that the algorithm is quite effective.

Incremental natural-gradient actor-critic algorithms

We present four new reinforcement learning algorithms based on actor-critic and natural-gradient ideas, and provide their convergence proofs. Actor-critic rein- forcement learning methods are online approximations to policy iteration in which the value-function parameters are estimated using temporal difference learning and the policy parameters are updated by stochastic gradient descent. Methods based on policy gradients in this way are of special interest because of their com- patibility with function approximation methods, which are needed to handle large or infinite state spaces. The use of temporal difference learning in this way is of interest because in many applications it dramatically reduces the variance of the gradient estimates. The use of the natural gradient is of interest because it can produce better conditioned parameterizations and has been shown to further re- duce variance in some cases. Our results extend prior two-timescale convergence results for actor-critic methods by Konda and Tsitsiklis by using temporal differ- ence learning in the actor and by incorporating natural gradients, and they extend prior empirical studies of natural actor-critic methods by Peters, Vijayakumar and Schaal by providing the first convergence proofs and the first fully incremental algorithms.

Efficient Algorithms for Linear Summed Error Structural SVMs

Structural Support Vector Machines (SSVMs) have become a popular tool in machine learning for predicting structured objects like parse trees, Part-of-Speech (POS) label sequences and image segments. Various efficient algorithmic techniques have been proposed for training SSVMs for large datasets. The typical SSVM formulation contains a regularizer term and a composite loss term. The loss term is usually composed of the Linear Maximum Error (LME) associated with the training examples. Other alternatives for the loss term are yet to be explored for SSVMs. We formulate a new SSVM with Linear Summed Error (LSE) loss term and propose efficient algorithms to train the new SSVM formulation using primal cutting-plane method and sequential dual coordinate descent method. Numerical experiments on benchmark datasets demonstrate that the sequential dual coordinate descent method is faster than the cutting-plane method and reaches the steady-state generalization performance faster. It is thus a useful alternative for training SSVMs when linear summed error is used.

Stochastic optimization for adaptive labor staffing in service systems

Service systems are labor intensive. Further, the workload tends to vary greatly with time. Adapting the staffing levels to the workloads in such systems is nontrivial due to a large number of parameters and operational variations, but crucial for business objectives such as minimal labor inventory. One of the central challenges is to optimize the staffing while maintaining system steady-state and compliance to aggregate SLA constraints. We formulate this problem as a parametrized constrained Markov process and propose a novel stochastic optimization algorithm for solving it. Our algorithm is a multi-timescale stochastic approximation scheme that incorporates a SPSA based algorithm for ‘primal descent' and couples it with a ‘dual ascent' scheme for the Lagrange multipliers. We validate this optimization scheme on five real-life service systems and compare it with a state-of-the-art optimization tool-kit OptQuest. Being two orders of magnitude faster than OptQuest, our scheme is particularly suitable for adaptive labor staffing. Also, we observe that it guarantees convergence and finds better solutions than OptQuest in many cases.