Clustering Based Large Margin Classification: A Scalable Approach using SOCP Formulation

This paper presents a novel Second Order Cone Programming (SOCP) formulation for large scale binary classification tasks. Assuming that the class conditional densities are mixture distributions, where each component of the mixture has a spherical covariance, the second order statistics of the components can be estimated efficiently using clustering algorithms like BIRCH. For each cluster, the second order moments are used to derive a second order cone constraint via a Chebyshev-Cantelli inequality. This constraint ensures that any data point in the cluster is classified correctly with a high probability. This leads to a large margin SOCP formulation whose size depends on the number of clusters rather than the number of training data points. Hence, the proposed formulation scales well for large datasets when compared to the state-of-the-art classifiers, Support Vector Machines (SVMs). Experiments on real world and synthetic datasets show that the proposed algorithm outperforms SVM solvers in terms of training time and achieves similar accuracies.

Reactive power optimization with different objectives in large power systems including HVDC systems and FACTS devices

Present day power systems are growing in size and complexity of operation with inter connections to neighboring systems, introduction of large generating units, EHV 400/765 kV AC transmission systems, HVDC systems and more sophisticated control devices such as FACTS. For planning and operational studies, it requires suitable modeling of all components in the power system, as the number of HVDC systems and FACTS devices of different type are incorporated in the system. This paper presents reactive power optimization with three objectives to minimize the sum of the squares of the voltage deviations (ve) of the load buses, minimization of sum of squares of voltage stability L-indices of load buses (¿L2), and also the system real power loss (Ploss) minimization. The proposed methods have been tested on typical sample system. Results for Indian 96-bus equivalent system including HVDC terminal and UPFC under normal and contingency conditions are presented.

Large improvements in application throughput of long-running multi-component applications using batch grids

Computational grids with multiple batch systems (batch grids) can be powerful infrastructures for executing long-running multi-component parallel applications. In this paper, we evaluate the potential improvements in throughput of long-running multi-component applications when the different components of the applications are executed on multiple batch systems of batch grids. We compare the multiple batch executions with executions of the components on a single batch system without increasing the number of processors used for executions. We perform our analysis with a foremost long-running multi-component application for climate modeling, the Community Climate System Model (CCSM). We have built a robust simulator that models the characteristics of both the multi-component application and the batch systems. By conducting large number of simulations with different workload characteristics and queuing policies of the systems, processor allocations to components of the application, distributions of the components to the batch systems and inter-cluster bandwidths, we show that multiple batch executions lead to 55% average increase in throughput over single batch executions for long-running CCSM. We also conducted real experiments with a practical middleware infrastructure and showed that multi-site executions lead to effective utilization of batch systems for executions of CCSM and give higher simulation throughput than single-site executions. Copyright (c) 2011 John Wiley & Sons, Ltd.

EXIT Chart Based Design of Irregular LDPC Codes for Large-MIMO Systems

In this letter, we characterize the extrinsic information transfer (EXIT) behavior of a factor graph based message passing algorithm for detection in large multiple-input multiple-output (MIMO) systems with tens to hundreds of antennas. The EXIT curves of a joint detection-decoding receiver are obtained for low density parity check (LDPC) codes of given degree distributions. From the obtained EXIT curves, an optimization of the LDPC code degree profiles is carried out to design irregular LDPC codes matched to the large-MIMO channel and joint message passing receiver. With low complexity joint detection-decoding, these codes are shown to perform better than off-the-shelf irregular codes in the literature by about 1 to 1.5 dB at a coded BER of 10(-5) in 16 x 16, 64 x 64 and 256 x 256 MIMO systems.

Scalable multi-dimensional user intent identification using tree structured distributions

The problem of identifying user intent has received considerable attention in recent years, particularly in the context of improving the search experience via query contextualization. Intent can be characterized by multiple dimensions, which are often not observed from query words alone. Accurate identification of Intent from query words remains a challenging problem primarily because it is extremely difficult to discover these dimensions. The problem is often significantly compounded due to lack of representative training sample. We present a generic, extensible framework for learning the multi-dimensional representation of user intent from the query words. The approach models the latent relationships between facets using tree structured distribution which leads to an efficient and convergent algorithm, FastQ, for identifying the multi-faceted intent of users based on just the query words. We also incorporated WordNet to extend the system capabilities to queries which contain words that do not appear in the training data. Empirical results show that FastQ yields accurate identification of intent when compared to a gold standard.

Linear Coding Schemes for the Distributed Computation of Subspaces

Let X-1,..., X-m be a set of m statistically dependent sources over the common alphabet F-q, that are linearly independent when considered as functions over the sample space. We consider a distributed function computation setting in which the receiver is interested in the lossless computation of the elements of an s-dimensional subspace W spanned by the elements of the row vector X-1,..., X-m]Gamma in which the (m x s) matrix Gamma has rank s. A sequence of three increasingly refined approaches is presented, all based on linear encoders. The first approach uses a common matrix to encode all the sources and a Korner-Marton like receiver to directly compute W. The second improves upon the first by showing that it is often more efficient to compute a carefully chosen superspace U of W. The superspace is identified by showing that the joint distribution of the {X-i} induces a unique decomposition of the set of all linear combinations of the {X-i}, into a chain of subspaces identified by a normalized measure of entropy. This subspace chain also suggests a third approach, one that employs nested codes. For any joint distribution of the {X-i} and any W, the sum-rate of the nested code approach is no larger than that under the Slepian-Wolf (SW) approach. Under the SW approach, W is computed by first recovering each of the {X-i}. For a large class of joint distributions and subspaces W, the nested code approach is shown to improve upon SW. Additionally, a class of source distributions and subspaces are identified, for which the nested-code approach is sum-rate optimal.

Probing the indefinite CP nature of the Higgs boson through decay distributions in the process e(+)e(-) -> t(t)over-bar Phi

The recently discovered scalar resonance at the Large Hadron Collider is now almost confirmed to be a Higgs boson, whose CP properties are yet to be established. At the International Linear Collider with and without polarized beams, it may be possible to probe these properties at high precision. In this work, we study the possibility of probing departures from the pure CP-even case, by using the decay distributions in the process e(+)e(-) -> t (t) over bar Phi, with Phi mainly decaying into a b (b) over bar pair. We have compared the case of a minimal extension of the Standard Model case (model I) with an additional pseudoscalar degree of freedom, with a more realistic case namely the CP-violating two-Higgs doublet model (model II) that permits a more general description of the couplings. We have considered the International Linear Collider with root s = 800 GeV and integrated luminosity of 300 fb(-1). Our main findings are that even in the case of small departures from the CP-even case, the decay distributions are sensitive to the presence of a CP-odd component in model II, while it is difficult to probe these departures in model I unless the pseudoscalar component is very large. Noting that the proposed degrees of beam polarization increase the statistics, the process demonstrates the effective role of beam polarization in studies beyond the Standard Model. Further, our study shows that an indefinite CP Higgs would be a sensitive laboratory to physics beyond the Standard Model.

Dynamics of prolate spheroidal mass distributions with varying eccentricity

In this paper we calculate the potential for a prolate spheroidal distribution as in a dark matter halo with a radially varying eccentricity. This is obtained by summing up the shell-by-shell contributions of isodensity surfaces, which are taken to be concentric and with a common polar axis and with an axis ratio that varies with radius. Interestingly, the constancy of potential inside a shell is shown to be a good approximation even when the isodensity contours are dissimilar spheroids, as long as the radial variation in eccentricity is small as seen in realistic systems. We consider three cases where the isodensity contours are more prolate at large radii, or are less prolate or have a constant eccentricity. Other relevant physical quantities like the rotation velocity, the net orbital and vertical frequency due to the halo and an exponential disc of finite thickness embedded in it are obtained. We apply this to the kinematical origin of Galactic warp, and show that a prolate-shaped halo is not conducive to making long-lived warps - contrary to what has been proposed in the literature. The results for a prolate mass distribution with a variable axis ratio obtained are general, and can be applied to other astrophysical systems, such as prolate bars, for a more realistic treatment.

Taming of large diameter triaxial setup

Triaxial tests are essential to estimate the shear strength properties of the soil or rock. Normally triaxial tests are carried out on samples of 38 mm diameter and 76 mm height. Granular materials, predominantly used in base/sub-base construction of pavements or in railways have size range of 60-75 mm. Determination of shear strength parameters of those materials can be made possible only through triaxial tests on large diameter samples. This paper describes a large diameter cyclic triaxial testing facility set up in the Geotechnical Engineering lab of Indian Institute of Science. This setup consists of 100 kN capacity dynamic loading frame, which facilitates testing of samples of up to 300 mm diameter and 600 mm height. The loading ram can be actuated up to a maximum frequency of 10 Hz, with maximum amplitude of 100 mm. The setup is capable of carrying out static as well as dynamic triaxial tests under isotropic, anisotropic conditions with a maximum confining pressure of 1 MPa. Working with this setup is a difficult task because of the size of the sample. In this paper, a detailed discussion on the various problems encountered during the initial testing using the equipment, the ideas and solutions adopted to solve them are presented. Pilot experiments on granular sub-base material of 53 mm down size are also presented.

Polymer template-assisted microemulsion synthesis of large surface area, porous Li2MnO3 and its characterization as a positive electrode material of Li-ion cells

Lithium-rich manganese oxide (Li2MnO3) is prepared by reverse microemulsion method employing Pluronic acid (P123) as a soft template and studied as a positive electrode material. The as-prepared sample possesses good crystalline structure with a broadly distributed mesoporosity but low surface area. As expected, cyclic voltammetry and charge-discharge data indicate poor electrochemical activity. However, the sample gains surface area with narrowly distributed mesoporosity and also electrochemical activity after treating in 4 M H2SO4. A discharge capacity of about 160 mAh g(-1) is obtained. When the acid-treated sample is heated at 300 A degrees C, the resulting porous sample with a large surface area and dual porosity provides a discharge capacity of 240 mAh g(-1). The rate capability study suggests that the sample provides about 150 mAh g(-1) at a specific discharge current of 1.25 A g(-1). Although the cycling stability is poor, the high rate capability is attributed to porous nature of the material.

Simulation of inhomogeneous distributions of ultracold atoms in an optical lattice via a massively parallel implementation of nonequilibrium strong-coupling perturbation theory

We present a nonequilibrium strong-coupling approach to inhomogeneous systems of ultracold atoms in optical lattices. We demonstrate its application to the Mott-insulating phase of a two-dimensional Fermi-Hubbard model in the presence of a trap potential. Since the theory is formulated self-consistently, the numerical implementation relies on a massively parallel evaluation of the self-energy and the Green's function at each lattice site, employing thousands of CPUs. While the computation of the self-energy is straightforward to parallelize, the evaluation of the Green's function requires the inversion of a large sparse 10(d) x 10(d) matrix, with d > 6. As a crucial ingredient, our solution heavily relies on the smallness of the hopping as compared to the interaction strength and yields a widely scalable realization of a rapidly converging iterative algorithm which evaluates all elements of the Green's function. Results are validated by comparing with the homogeneous case via the local-density approximation. These calculations also show that the local-density approximation is valid in nonequilibrium setups without mass transport.

DEEP WASHINGTON PHOTOMETRY OF INCONSPICUOUS STAR CLUSTER CANDIDATES IN THE LARGE MAGELLANIC CLOUD

We present deep Washington photometry of 45 poorly populated star cluster candidates in the Large Magellanic Cloud (LMC). We have performed a systematic study to estimate the parameters of the cluster candidates by matching theoretical isochrones to the cleaned and dereddened cluster color-magnitude diagrams. We were able to estimate the basic parameters for 33 clusters, out of which 23 are identified as single clusters and 10 are found to be members of double clusters. The other 12 cluster candidates have been classified as possible clusters/asterisms. About 50% of the true clusters are in the 100-300 Myr age range, whereas some are older or younger. We have discussed the distribution of age, location, and reddening with respect to field, as well as the size of true clusters. The sizes and masses of the studied sample are found to be similar to that of open clusters in the Milky Way. Our study adds to the lower end of cluster mass distribution in the LMC, suggesting that the LMC, apart from hosting rich clusters, also has formed small, less massive open clusters in the 100-300 Myr age range.

Efficient Dictionary for Salted Password Analysis

User authentication is essential for accessing computing resources, network resources, email accounts, online portals etc. To authenticate a user, system stores user credentials (user id and password pair) in system. It has been an interested field problem to discover user password from a system and similarly protecting them against any such possible attack. In this work we show that passwords are still vulnerable to hash chain based and efficient dictionary attacks. Human generated passwords use some identifiable patterns. We have analysed a sample of 19 million passwords, of different lengths, available online and studied the distribution of the symbols in the password strings. We show that the distribution of symbols in user passwords is affected by the native language of the user. From symbol distributions we can build smart and efficient dictionaries, which are smaller in size and their coverage of plausible passwords from Key-space is large. These smart dictionaries make dictionary based attacks practical.

Wireless scheduling with partial channel state information: large deviations and optimality

We consider a server serving a time-slotted queued system of multiple packet-based flows, where not more than one flow can be serviced in a single time slot. The flows have exogenous packet arrivals and time-varying service rates. At each time, the server can observe instantaneous service rates for only a subset of flows ( selected from a fixed collection of observable subsets) before scheduling a flow in the subset for service. We are interested in queue length aware scheduling to keep the queues short. The limited availability of instantaneous service rate information requires the scheduler to make a careful choice of which subset of service rates to sample. We develop scheduling algorithms that use only partial service rate information from subsets of channels, and that minimize the likelihood of queue overflow in the system. Specifically, we present a new joint subset-sampling and scheduling algorithm called Max-Exp that uses only the current queue lengths to pick a subset of flows, and subsequently schedules a flow using the Exponential rule. When the collection of observable subsets is disjoint, we show that Max-Exp achieves the best exponential decay rate, among all scheduling algorithms that base their decision on the current ( or any finite past history of) system state, of the tail of the longest queue. To accomplish this, we employ novel analytical techniques for studying the performance of scheduling algorithms using partial state, which may be of independent interest. These include new sample-path large deviations results for processes obtained by non-random, predictable sampling of sequences of independent and identically distributed random variables. A consequence of these results is that scheduling with partial state information yields a rate function significantly different from scheduling with full channel information. In the special case when the observable subsets are singleton flows, i.e., when there is effectively no a priori channel state information, Max-Exp reduces to simply serving the flow with the longest queue; thus, our results show that to always serve the longest queue in the absence of any channel state information is large deviations optimal.

Handling large variations in mechanics: Some applications

There is a need to use probability distributions with power-law decaying tails to describe the large variations exhibited by some of the physical phenomena. The Weierstrass Random Walk (WRW) shows promise for modeling such phenomena. The theory of anomalous diffusion is now well established. It has found number of applications in Physics, Chemistry and Biology. However, its applications are limited in structural mechanics in general, and structural engineering in particular. The aim of this paper is to present some mathematical preliminaries related to WRW that would help in possible applications. In the limiting case, it represents a diffusion process whose evolution is governed by a fractional partial differential equation. Three applications of superdiffusion processes in mechanics, illustrating their effectiveness in handling large variations, are presented.