Scalable non-linear Support Vector Machine using hierarchical clustering

This paper discusses a method for scaling SVM with Gaussian kernel function to handle large data sets by using a selective sampling strategy for the training set. It employs a scalable hierarchical clustering algorithm to construct cluster indexing structures of the training data in the kernel induced feature space. These are then used for selective sampling of the training data for SVM to impart scalability to the training process. Empirical studies made on real world data sets show that the proposed strategy performs well on large data sets.

Towards building an intelligent decision support system for project management

Management of large projects, especially the ones in which a major component of R&D is involved and those requiring knowledge from diverse specialised and sophisticated fields, may be classified as semi-structured problems. In these problems, there is some knowledge about the nature of the work involved, but there are also uncertainties associated with emerging technologies. In order to draw up a plan and schedule of activities of such a large and complex project, the project manager is faced with a host of complex decisions that he has to take, such as, when to start an activity, for how long the activity is likely to continue, etc. An Intelligent Decision Support System (IDSS) which aids the manager in decision making and drawing up a feasible schedule of activities while taking into consideration the constraints of resources and time, will have a considerable impact on the efficient management of the project. This report discusses the design of an IDSS that helps in project planning phase through the scheduling phase. The IDSS uses a new project scheduling tool, the Project Influence Graph (PIG).

Believe it or Not! Multicore CPUs can Match GPUs for FLOP-intensive Applications

In this work, we evaluate performance of a real-world image processing application that uses a cross-correlation algorithm to compare a given image with a reference one. The algorithm processes individual images represented as 2-dimensional matrices of single-precision floating-point values using O(n4) operations involving dot-products and additions. We implement this algorithm on a nVidia GTX 285 GPU using CUDA, and also parallelize it for the Intel Xeon (Nehalem) and IBM Power7 processors, using both manual and automatic techniques. Pthreads and OpenMP with SSE and VSX vector intrinsics are used for the manually parallelized version, while a state-of-the-art optimization framework based on the polyhedral model is used for automatic compiler parallelization and optimization. The performance of this algorithm on the nVidia GPU suffers from: (1) a smaller shared memory, (2) unaligned device memory access patterns, (3) expensive atomic operations, and (4) weaker single-thread performance. On commodity multi-core processors, the application dataset is small enough to fit in caches, and when parallelized using a combination of task and short-vector data parallelism (via SSE/VSX) or through fully automatic optimization from the compiler, the application matches or beats the performance of the GPU version. The primary reasons for better multi-core performance include larger and faster caches, higher clock frequency, higher on-chip memory bandwidth, and better compiler optimization and support for parallelization. The best performing versions on the Power7, Nehalem, and GTX 285 run in 1.02s, 1.82s, and 1.75s, respectively. These results conclusively demonstrate that, under certain conditions, it is possible for a FLOP-intensive structured application running on a multi-core processor to match or even beat the performance of an equivalent GPU version.

Method for Creative Behavioral Design Based on Analogy and Blending from Natural Things

In the near future, robots and CG (computer graphics) will be required to exhibit creative behaviors that reflect designers’ abstract images and emotions. However, there are no effective methods to develop abstract images and emotions and support designers in designing creative behaviors that reflect their images and emotions. Analogy and blending are two methods known to be very effective for designing creative behaviors. The aim of this study is to propose a method for developing designers’ abstract behavioral images and emotions and giving shape to them by constructing a computer system that supports a designer in the creation of the desired behavior. This method focuses on deriving inspiration from the behavioral aspects of natural phenomena rather than simply mimicking it. We have proposed two new methods for developing abstract behavioral images and emotions by which a designer can use analogies from natural things such as animals and plants even when there is a difference in the number of joints between the natural object and the design target. The first method uses visual behavioral images, the second uses rhythmic behavioral images. We have demonstrated examples of designed behaviors to verify the effectiveness of the proposed methods.

Geographic Resources Decision Support System – an Open Source GIS

Fully structured and matured open source spatial and temporal analysis technology seems to be the official carrier of the future for planning of the natural resources especially in the developing nations. This technology has gained enormous momentum because of technical superiority, affordability and ability to join expertise from all sections of the society. Sustainable development of a region depends on the integrated planning approaches adopted in decision making which requires timely and accurate spatial data. With the increased developmental programmes, the need for appropriate decision support system has increased in order to analyse and visualise the decisions associated with spatial and temporal aspects of natural resources. In this regard Geographic Information System (GIS) along with remote sensing data support the applications that involve spatial and temporal analysis on digital thematic maps and the remotely sensed images. Open source GIS would help in wide scale applications involving decisions at various hierarchical levels (for example from village panchayat to planning commission) on economic viability, social acceptance apart from technical feasibility. GRASS (Geographic Resources Analysis Support System, http://wgbis.ces.iisc.ernet.in/grass) is an open source GIS that works on Linux platform (freeware), but most of the applications are in command line argument, necessitating a user friendly and cost effective graphical user interface (GUI). Keeping these aspects in mind, Geographic Resources Decision Support System (GRDSS) has been developed with functionality such as raster, topological vector, image processing, statistical analysis, geographical analysis, graphics production, etc. This operates through a GUI developed in Tcltk (Tool command language / Tool kit) under Linux as well as with a shell in X-Windows. GRDSS include options such as Import /Export of different data formats, Display, Digital Image processing, Map editing, Raster Analysis, Vector Analysis, Point Analysis, Spatial Query, which are required for regional planning such as watershed Analysis, Landscape Analysis etc. This is customised to Indian context with an option to extract individual band from the IRS (Indian Remote Sensing Satellites) data, which is in BIL (Band Interleaved by Lines) format. The integration of PostgreSQL (a freeware) in GRDSS aids as an efficient database management system.

A Randomized Algorithm for Large Scale Support Vector Learning

We propose a randomized algorithm for large scale SVM learning which solves the problem by iterating over random subsets of the data. Crucial to the algorithm for scalability is the size of the subsets chosen. In the context of text classification we show that, by using ideas from random projections, a sample size of O(log n) can be used to obtain a solution which is close to the optimal with a high probability. Experiments done on synthetic and real life data sets demonstrate that the algorithm scales up SVM learners, without loss in accuracy. 1

Decision support system for regional electricity planning

Electricity appears to be the energy carrier of choice for modern economics since growth in electricity has outpaced growth in the demand for fuels. A decision maker (DM) for accurate and efficient decisions in electricity distribution requires the sector wise and location wise electricity consumption information to predict the requirement of electricity. In this regard, an interactive computer-based Decision Support System (DSS) has been developed to compile, analyse and present the data at disaggregated levels for regional energy planning. This helps in providing the precise information needed to make timely decisions related to transmission and distribution planning leading to increased efficiency and productivity. This paper discusses the design and implementation of a DSS, which facilitates to analyse the consumption of electricity at various hierarchical levels (division, taluk, sub division, feeder) for selected periods. This DSS is validated with the data of transmission and distribution systems of Kolar district in Karnataka State, India.

Knowledge-Based Approach Using Support Vector Machine for Transmission Line Distance Relay Co-ordination

In this paper, knowledge-based approach using Support Vector Machines (SVMs) are used for estimating the coordinated zonal settings of a distance relay. The approach depends on the detailed simulation studies of apparent impedance loci as seen by distance relay during disturbance, considering various operating conditions including fault resistance. In a distance relay, the impedance loci given at the relay location is obtained from extensive transient stability studies. SVMs are used as a pattern classifier for obtaining distance relay co-ordination. The scheme utilizes the apparent impedance values observed during a fault as inputs. An improved performance with the use of SVMs, keeping the reach when faced with different fault conditions as well as system power flow changes, are illustrated with an equivalent 265 bus system of a practical Indian Western Grid.

Infinite push: a new support vector ranking algorithm that directly optimizes accuracy at the absolute top of the list

Ranking problems have become increasingly important in machine learning and data mining in recent years, with applications ranging from information retrieval and recommender systems to computational biology and drug discovery. In this paper, we describe a new ranking algorithm that directly maximizes the number of relevant objects retrieved at the absolute top of the list. The algorithm is a support vector style algorithm, but due to the different objective, it no longer leads to a quadratic programming problem. Instead, the dual optimization problem involves l1, ∞ constraints; we solve this dual problem using the recent l1, ∞ projection method of Quattoni et al (2009). Our algorithm can be viewed as an l∞-norm extreme of the lp-norm based algorithm of Rudin (2009) (albeit in a support vector setting rather than a boosting setting); thus we refer to the algorithm as the ‘Infinite Push’. Experiments on real-world data sets confirm the algorithm’s focus on accuracy at the absolute top of the list.

A study on the usefulness of Support Vector Machines for the realtime computational simulation of soft biological organs

Realistic and realtime computational simulation of soft biological organs (e.g., liver, kidney) is necessary when one tries to build a quality surgical simulator that can simulate surgical procedures involving these organs. Since the realistic simulation of these soft biological organs should account for both nonlinear material behavior and large deformation, achieving realistic simulations in realtime using continuum mechanics based numerical techniques necessitates the use of a supercomputer or a high end computer cluster which are costly. Hence there is a need to employ soft computing techniques like Support Vector Machines (SVMs) which can do function approximation, and hence could achieve physically realistic simulations in realtime by making use of just a desktop computer. Present work tries to simulate a pig liver in realtime. Liver is assumed to be homogeneous, isotropic, and hyperelastic. Hyperelastic material constants are taken from the literature. An SVM is employed to achieve realistic simulations in realtime, using just a desktop computer. The code for the SVM is obtained from [1]. The SVM is trained using the dataset generated by performing hyperelastic analyses on the liver geometry, using the commercial finite element software package ANSYS. The methodology followed in the present work closely follows the one followed in [2] except that [2] uses Artificial Neural Networks (ANNs) while the present work uses SVMs to achieve realistic simulations in realtime. Results indicate the speed and accuracy that is obtained by employing the SVM for the targeted realistic and realtime simulation of the liver.

External support and innovation performance of SMEs in Bangalore: Role of firm level factors

This paper probes the role of internal factors in SMEs in obtaining external support and achieving innovation performance in the context of auto component, electronics and machine tool industries of Bangalore in India. Using step-wise logistic regression analysis, the study found that only if SMEs have internal technical competence in terms of technically qualified entrepreneur, an exclusive design centre, and innovate more frequently, they will be able to obtain external support. Further using step-wise multiple regression the study concluded that SMEs which have come up to implement innovative ideas or exploit market opportunities and which have obtained external support with technically qualified entrepreneurs are able to exhibit better innovation performance.