Bhubaneshwar, India
Bhubaneshwar, India

Time filter

Source Type

Mohapatra P.,IIIT | Chakravarty S.,Orissa Engineering College, Bhubaneswar | Dash P.K.,Siksha ‘O’ Anusandhan University
Swarm and Evolutionary Computation | Year: 2016

Microarray gene expression based medical data classification has remained as one of the most challenging research areas in the field of bioinformatics, machine learning and pattern classification. This paper proposes two variations of kernel ridge regression (KRR), namely wavelet kernel ridge regression (WKRR) and radial basis kernel ridge regression (RKRR) for classification of microarray medical datasets. Microarray medical datasets contain irrelevant and redundant genes which cause high number of gene expression i.e. dimensionality and small sample sizes. To overcome the curse of dimensionality of the microarray datasets, modified cat swarm optimization (MCSO), a naturally inspired evolutionary algorithm, is used to select the most relevant features from the datasets. The adequacies of the classifiers are demonstrated by employing four from each binary and multi-class microarray medical datasets. Breast cancer, prostate cancer, colon tumor, leukemia datasets belong to the former and leukemia1, leukemia2, SRBCT, brain tumor1 to the latter. A number of useful performance evaluation measures including accuracy, sensitivity, specificity, confusion matrix, Gmean, F-score and the area under the receiver operating characteristic (ROC) curve are considered to examine the efficacy of the model. Other models like simple ridge regression (RR), online sequential ridge regression (OSRR), support vector machine radial basis function (SVMRBF), support vector machine polynomial (SVMPoly) and random forest are studied and analyzed for comparison. The experimental results demonstrate that KRR outperforms other models irrespective of the datasets and WKRR produces better results as compared to RKRR. Finally, when the results are compared on the basis of binary and multi-class datasets, it is found that binary class yields a little bit better result as compared to the multiclass irrespective of models. © 2016 Elsevier B.V.

Mohapatra P.,IIIT | Chakravarty S.,Orissa Engineering College, Bhubaneswar | Dash P.K.,Siksha ‘O’ Anusandhan University
Swarm and Evolutionary Computation | Year: 2015

Machine learning techniques are being increasingly used for detection and diagnosis of diseases for its accuracy and efficiency in pattern classification. In this paper, improved cuckoo search based extreme learning machine (ICSELM) is proposed to classify binary medical datasets. Extreme learning machine (ELM) is widely used as a learning algorithm for training single layer feed forward neural networks (SLFN) in the field of classification. However, to make the model more stable, an evolutionary algorithm improved cuckoo search (ICS) is used to pre-train ELM by selecting the input weights and hidden biases. Like ELM, Moore-Penrose (MP) generalized inverse is used in ICSELM to analytically determines the output weights. To evaluate the effectiveness of the proposed model, four benchmark datasets, i.e. Breast Cancer, Diabetes, Bupa and Hepatitis from the UCI Repository of Machine Learning are used. A number of useful performance evaluation measures including accuracy, sensitivity, specificity, confusion matrix, Gmean, F-score and norm of the output weights as well as the area under the receiver operating characteristic (ROC) curve are computed. The results are analyzed and compared with both ELM based models like ELM, on-line sequential extreme learning algorithm (OSELM), CSELM and other artificial neural networks i.e. multi-layered perceptron (MLP), MLPCS, MLPICS and radial basis function neural network (RBFNN), RBFNNCS, RBFNNICS. The experimental results demonstrate that the ICSELM model outperforms other models. © 2015 Elsevier B.V.

Kar A.,IIIT | Chandra M.,BIT Mesra
International Journal of Computational Vision and Robotics | Year: 2013

The structural complexity and overall performance of the adaptive filter depend on its structure. The number of taps is one of the most important structural parameters of the liner adaptive filter. In practice the system length is not known a priori and has to be estimated from the knowledge of the input and output signals. In a system identification framework the tap-length estimation algorithm automatically adapts the filter order to the suitable optimum value which makes the variable order adaptive filter a best identifier of the unknown plant. In this paper an improved pseudo-fractional tap-length selection algorithm is proposed to find out the optimum tap-length which best balances the complexity and steady state performance. The performance analysis is presented to formulate steady state tap-length in correspondence with the proposed algorithm. Simulations and results are provided to observe the analysis and to make a comparison with the existing tap-length learning methods. Copyright © 2013 Inderscience Enterprises Ltd.

Sinha N.,IIIT | Babu R.V.,Indian Institute of Science
Proceedings - International Conference on Image Processing, ICIP | Year: 2012

Automatic eye screening for conditions like diabetic retinopathy critically hinges on detection and localization of Optic disk (OD). In this paper, we present a novel scale-embedded dictionary-based method that poses the problem of OD localization as that of classification, carried out in sparse representation framework. A dictionary is created with manually marked fixed-sized sub-images that contain OD at the center, for multiple scales. For a given test image, all subimages are sparsely represented as a linear combination of OD dictionary elements. A confidence measure indicating the likelihood of the presence of OD is obtained from these coefficients. Red channel and gray intensity images are processed independently, and their respective confidence measures are fused to form a confidence map. A blob detector is run on the confidence map, whose peak response is considered to be at the location of the OD. The proposed method is evaluated on publicly available databases such as DIARETDB0, DIARETDB1 and DRIVE. The OD was correctly localized in 253 out of 259 images, with an average computation time of 3.8 seconds/image and accuracy of 97.6%. Comparisons with two existing techniques are also discussed. © 2012 IEEE.

The structure of an adaptive time varying linear filter largely depends on its tap-length and the delay units connected to it. The no of taps is one of the most important structural parameters of the liner adaptive filter. Determining the system order or length is not a trivial task. Fixing the tap-length at a fixed value sometimes results in unavoidable issues with the adaptive design like insufficient modeling and adaptation noise. On the other hand a dynamic tap-length adaptation algorithm automatically finds the optimum order of the adaptive filter to have a tradeoff between the convergence and steady state error. It is always difficult to get satisfactory performance in high noise environment employing an adaptive filter for any identification problem. High noise decreases the Signal to noise ratio and sometimes creates wandering issues. In this chapter an improved pseudo-fractional tap-length selection algorithm has been proposed and analyzed to find out the optimum tap-length which best balances the complexity and steady state performance specifically in high noise environment. A steady-state performance analysis has been presented to formulate the steady state tap-length in correspondence with the proposed algorithm. Simulations and results are provided to observe the analysis and to make a comparison with the existing tap-length learning methods. © Springer International Publishing Switzerland 2014.

Bose S.K.,Unisys Corporation | Brock S.,Unisys Corporation | Skeoch R.,Unisys Corporation | Rao S.,IIIT
Proceedings - 11th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2011 | Year: 2011

A virtual machine (VM), during its lifecycle, can be scheduled for execution at geographically disparate cloud locations depending upon the cost of computation and the load at these locations. However, trans-locating a live VM across high-latency low-bandwidth wide area networks (WAN) within 'reasonable' time is nearly impossible due to the large size of the VM image. In this paper, we deal with this problem by combining VM scheduling strategies with VM replication strategies. In particular, we propose to replicate a VM image selectively across different cloud sites, choose a replica of the VM image to be the primary copy and propagate the incremental changes at the primary copy to all the remaining replicas of the VM image. The replica placement strategies are based on factors that influence long-term costs such as the average per-unit cost of storage and the average per-unit cost of computation at different cloud sites besides the 'end-user' latency requirements associated with the VMs. We propose to compensate the additional storage requirements due to replication by exploring commonality that naturally exists amongst different VM images using de-duplication techniques. A key issue that naturally arises in this integrated replication and scheduling context for minimizing migration latencies associated with live migration of VMs across WAN is the design of a good replica placement algorithm that minimizes additional storage requirements. In this paper we address this issue as part for our integrated replication and scheduling architecture, called Cloud Spider. We discuss the trade-offs involved in the design of a replica placement algorithm and propose an algorithm that factors in deduplication ratios amongst pairs of VM images while deciding on the question of replica placement of the VM images. Preliminary experiments show extremely promising results. © 2011 IEEE.

Sahoo H.K.,IIIT | Dash P.K.,Siksha ‘O’ Anusandhan University | Rath N.P.,VSSUT
AEU - International Journal of Electronics and Communications | Year: 2012

Mechanical vibration signals are always composed of harmonics of different order. A novel estimator is proposed for estimating the frequency of sinusoidal signals from measurements corrupted by White Gaussian noise with zero mean. Also low frequency sinusoidal signal is considered along with third and fifth order harmonics in presence of noise for estimating amplitudes and phases of different harmonics. The proposed estimator known as complex H∞ filter is applied to a noisy sinusoidal signal model. State space modeling with two and three state variables is used for estimation of frequency in presence of white noise. Various comparisons in terms of simulation results for time varying frequency reveal that the proposed adaptive filter has significant improvement in noise rejection and estimation accuracy. Comparison in performance between two and three states modeling is presented in terms of mean square error (MSE) under different SNR conditions.The computer simulations clearly indicate that two states modeling based on Hilbert transform performs better than three states modeling under high noisy condition. Frequency estimation performance of the proposed filter is also being compared with extended complex Kalman filter (ECKF) under same noisy conditions through simulations. © 2011 Elsevier GmbH. All rights reserved.

Singh V.P.,Babu Banarsi das National Institute of Technology and Management | Mohanty S.R.,Motilal Nehru National Institute of Technology | Kishor N.,Motilal Nehru National Institute of Technology | Ray P.K.,IIIT
International Journal of Electrical Power and Energy Systems | Year: 2013

This paper presents a study on isolated hybrid distributed generation (DG) system for improving the frequency deviation profile. The hybrid DG system consists of wind turbine generator (WTG), diesel engine generator (DEG), aqua-electrolyzer (AE), fuel cell (FC) along with energy storage units. The frequency control problem is addressed for DG system connected with superconducting magnetic energy storage (SMES) or ultra-capacitor (UC). The particle swarm optimization (PSO) based loop shaping of H-infinity controller is used and compared with those obtained by genetic algorithm (GA) to minimize the frequency deviation. The frequency stabilizing performance is analyzed under different disturbances. Also, the controller robustness in terms of system parameter uncertainties is tested for changes in parameter up to ±30% from its nominal value. The results demonstrate minimum frequency deviation as achieved by proposed controller with use of UC in hybrid DG system. © 2012 Elsevier Ltd. All rights reserved.

Saikia G.,IIIT
2014 International Conference on Signal Processing and Integrated Networks, SPIN 2014 | Year: 2014

Functionalities of the hand of the iCub simulator have been extended so as to make the simulator robot make grasp forms like a human hand. This is achieved by sending signal for one of the six common types of human hand grasps, viz. power, pinch, precision, hook, oblique and palm-up from a server to a kinematic solver which then calculates the joint angles for each finger of the robot and feeds these results to a motion controller which finally moves the finger actuators to form the grasp. A geometric approach to solve the manipulator kinematics is used owing to almost planar nature of the finger manipulators of the iCub hand and small number of independent joint variables. This © 2014 IEEE.

Sahoo H.K.,IIIT | Dash P.K.,Soa University | Rath N.P.,VSSUT
Applied Soft Computing Journal | Year: 2013

This paper proposes NARX (nonlinear autoregressive model with exogenous input) model structures with functional expansion of input patterns by using low complexity ANN (artificial neural network) for nonlinear system identification. Chebyshev polynomials, Legendre polynomials, trigonometric expansions using sine and cosine functions as well as wavelet basis functions are used for the functional expansion of input patterns. The past input and output samples are modeled as a nonlinear NARX process and robust H∞ filter is proposed as the learning algorithm for the neural network to identify the unknown plants. H∞ filtering approach is based on the state space modeling of model parameters and evaluation of Jacobian matrices. This approach is the robustification of Kalman filter which exhibits robust characteristics and fast convergence properties. Comparison results for different nonlinear dynamic plants with forgetting factor recursive least square (FFRLS) and extended Kalman filter (EKF) algorithms demonstrate the effectiveness of the proposed approach. © 2013 Elsevier B.V. All rights reserved.

Loading IIIT collaborators
Loading IIIT collaborators