Searcharxiv⌕ Search

arXiv subjects

Suresh Chandra

Publications and source records attributed to Suresh Chandra.

At least 19 recordsLinked to original sources
arXiv2025

GBSVR: Granular Ball Support Vector Regression

Support Vector Regression (SVR) and its variants are widely used to handle regression tasks, however, since their solution involves solving an expensive quadratic programming problem, it limits its application, especially when dealing with large datasets. Additionally, SVR uses an epsilon-insensitive loss function which is sensitive to outliers and therefore can adversely affect its performance. We propose Granular Ball Support Vector Regression (GBSVR) to tackle problem of regression by using granular ball concept. These balls are useful in simplifying complex data spaces for machine learning tasks, however, to the best of our knowledge, they have not been sufficiently explored for regression problems. Granular balls group the data points into balls based on their proximity and reduce the computational cost in SVR by replacing the large number of data points with far fewer granular balls. This work also suggests a discretization method for continuous-valued attributes to facilitate the construction of granular balls. The effectiveness of the proposed approach is evaluated on several benchmark datasets and it outperforms existing state-of-the-art approaches

cs.LG↗
arXiv2024

Tube Loss: A Novel Approach for Prediction Interval Estimation

This paper proposes a novel loss function, called 'Tube Loss', for simultaneous estimation of bounds of a Prediction Interval (PI) in the regression setup. The PIs obtained by minimizing the empirical risk based on the Tube Loss are shown to be of better quality than the PIs obtained by the existing methods in the following sense. First, it yields intervals that attain the prespecified confidence level t $\in$ (0,1) asymptotically. A theoretical proof of this fact is given. Secondly, the user is allowed to move the interval up or down by controlling the value of a parameter. This helps the user to choose a PI capturing denser regions of the probability distribution of the response variable inside the interval, and thus, sharpening its width. This is shown to be especially useful when the conditional distribution of the response variable is skewed. Further, the Tube Loss based PI estimation method can trade-off between the coverage and the average width by solving a single optimization problem. It enables further reduction of the average width of PI through re-calibration. Also, unlike a few existing PI estimation methods the gradient descent (GD) method can be used for minimization of empirical risk. Through extensive experiments, we demonstrate the effectiveness of Tube Loss-based PI estimation in both kernel machines and neural networks. Additionally, we show that Tube Loss-based deep probabilistic forecasting models achieve superior performance compared to existing probabilistic forecasting techniques across several benchmark and wind datasets. Finally, we empirically validate the advantages of the Tube loss approach within the conformal prediction framework. Codes are available at https://github.com/ltpritamanand/Tube$\_$loss.

cs.LG↗
arXiv2021

Improvement over Pinball Loss Support Vector Machine

Recently, there have been several papers that discuss the extension of the Pinball loss Support Vector Machine (Pin-SVM) model, originally proposed by Huang et al.,[1][2]. Pin-SVM classifier deals with the pinball loss function, which has been defined in terms of the parameter $τ$. The parameter $τ$ can take values in $[ -1,1]$. The existing Pin-SVM model requires to solve the same optimization problem for all values of $τ$ in $[ -1,1]$. In this paper, we improve the existing Pin-SVM model for the binary classification task. At first, we note that there is major difficulty in Pin-SVM model (Huang et al. [1]) for $ -1 \leq τ< 0$. Specifically, we show that the Pin-SVM model requires the solution of different optimization problem for $ -1 \leq τ< 0$. We further propose a unified model termed as Unified Pin-SVM which results in a QPP valid for all $-1\leq τ\leq 1$ and hence more convenient to use. The proposed Unified Pin-SVM model can obtain a significant improvement in accuracy over the existing Pin-SVM model which has also been empirically justified by extensive numerical experiments with real-world datasets.

cs.LG↗
arXiv2021

Twin Augmented Architectures for Robust Classification of COVID-19 Chest X-Ray Images

The gold standard for COVID-19 is RT-PCR, testing facilities for which are limited and not always optimally distributed. Test results are delayed, which impacts treatment. Expert radiologists, one of whom is a co-author, are able to diagnose COVID-19 positivity from Chest X-Rays (CXR) and CT scans, that can facilitate timely treatment. Such diagnosis is particularly valuable in locations lacking radiologists with sufficient expertise and familiarity with COVID-19 patients. This paper has two contributions. One, we analyse literature on CXR based COVID-19 diagnosis. We show that popular choices of dataset selection suffer from data homogeneity, leading to misleading results. We compile and analyse a viable benchmark dataset from multiple existing heterogeneous sources. Such a benchmark is important for realistically testing models. Our second contribution relates to learning from imbalanced data. Datasets for COVID X-Ray classification face severe class imbalance, since most subjects are COVID -ve. Twin Support Vector Machines (Twin SVM) and Twin Neural Networks (Twin NN) have, in recent years, emerged as effective ways of handling skewed data. We introduce a state-of-the-art technique, termed as Twin Augmentation, for modifying popular pre-trained deep learning models. Twin Augmentation boosts the performance of a pre-trained deep neural network without requiring re-training. Experiments show, that across a multitude of classifiers, Twin Augmentation is very effective in boosting the performance of given pre-trained model for classification in imbalanced settings.

cs.CV↗
arXiv2020

Support Vector Regression via a Combined Reward Cum Penalty Loss Function

In this paper, we introduce a novel combined reward cum penalty loss function to handle the regression problem. The proposed combined reward cum penalty loss function penalizes the data points which lie outside the $ε$-tube of the regressor and also assigns reward for the data points which lie inside of the $ε$-tube of the regressor. The combined reward cum penalty loss function based regression (RP-$ε$-SVR) model has several interesting properties which are investigated in this paper and are also supported with the experimental results.

cs.LG↗
arXiv2019

A $ν$- support vector quantile regression model with automatic accuracy control

This paper proposes a novel '$ν$-support vector quantile regression' ($ν$-SVQR) model for the quantile estimation. It can facilitate the automatic control over accuracy by creating a suitable asymmetric $ε$-insensitive zone according to the variance present in data. The proposed $ν$-SVQR model uses the $ν$ fraction of training data points for the estimation of the quantiles. In the $ν$-SVQR model, training points asymptotically appear above and below of the asymmetric $ε$-insensitive tube in the ratio of $1-τ$ and $τ$. Further, there are other interesting properties of the proposed $ν$-SVQR model, which we have briefly described in this paper. These properties have been empirically verified using the artificial and real world dataset also.

cs.LG↗
arXiv2019

A new asymmetric $ε$-insensitive pinball loss function based support vector quantile regression model

In this paper, we propose a novel asymmetric $ε$-insensitive pinball loss function for quantile estimation. There exists some pinball loss functions which attempt to incorporate the $ε$-insensitive zone approach in it but, they fail to extend the $ε$-insensitive approach for quantile estimation in true sense. The proposed asymmetric $ε$-insensitive pinball loss function can make an asymmetric $ε$- insensitive zone of fixed width around the data and divide it using $τ$ value for the estimation of the $τ$th quantile. The use of the proposed asymmetric $ε$-insensitive pinball loss function in Support Vector Quantile Regression (SVQR) model improves its prediction ability significantly. It also brings the sparsity back in SVQR model. Further, the numerical results obtained by several experiments carried on artificial and real world datasets empirically show the efficacy of the proposed `$ε$-Support Vector Quantile Regression' ($ε$-SVQR) model over other existing SVQR models.

stat.ML↗
arXiv2015

On the basis-set extrapolation

A possible solution for the problem of memory-size and computer-time, is the extrapolation of basis-set$^1$. This extrapolation has two exponents $α$ and $β$, corresponding to the HF (reference energy) and the energy of correlations (EC), respectively. For a given system, the exponents are taken as constant$^2$, and potential energy surfaces (PES) are generated. We have found that the values of $α$ and $β$ are not constant, but vary from position to position in the system. How to deal with such situation and get very accurate PES, is discussed.

physics.chem-ph↗
arXiv2015

Detection of TiH$_2$ molecule in the interstellar medium is less probable

Identification of TiH$^1$ and TiO$^2$ has been historical, as the Titanium was first time discovered in the interstellar medium (ISM). After finding TiO$_2$$^3$, there is an obvious question about the search of titanium dihydride (TiH$_2$). The existence of TiH$_2$ in the ISM is quite probable, as the atomic abundance of hydrogen is about 1900 times larger than that of oxygen. We have discussed that the detection of TiH$_2$ in the ISM is less probable, though it has a large electric dipole moment.

astro-ph.GA↗
arXiv2015

On partition function in Astronomy \& Astrophysics

In order to analyze spectrum from the interstellar medium (ISM), spectrum of the molecule of interest is recorded in a laboratory, and accurate rotational and centrifugal distortion constants are derived. By using these constants, one can calculate accurate partition function. However, in the same paper, where these constants are derived, the partition function is calculated by using a semi-empirical expression. We have looked into the details of this semi-empirical expression and compared the values, obtained from it, with the accurate ones. As an example, we have considered the case of Methanimine (CH$_2$NH) which is detected in a number of cosmic objects. It is found that for the kinetic temperature $T > 120$ K, the semi-empirical expression gives large value as compared to the accurate one. The deviation becomes about 25\% larger than the accurate one at the kinetic temperature of 400 K.

astro-ph.GA↗
arXiv2015

Earth Similarity Index with two free parameters

We have derived Earth Similarity Index (ESI) with two free parameters m and T. These free parameters are optimized with the consideration that the planet Mars is almost similar to the Earth. For the optimized values of free parameters, the interior-ESI, surface-ESI and ESI for some planets are calculated. The results for m = 0.8 and T = 0.8 are compared with the values obtained by Schulze-Makuch {\it et al.} (2011). We have found that the exoplanet 55 Cnc f is within 10% away from the threshold value T. The exoplanets HD 69830 c, 55 Cnc c, 55 Cnc f, 61 Vir d and HIP 57050 b are found to have ESI within 10% from the threshold value.

astro-ph.EP↗
arXiv2015

Some observations about the MOLSCAT

For calculation of cross sections for collisional transitions between rotational levels in a molecule, a computer code, MOLSCAT has been developed by Hutson \& Green (1994). For the transitions between rotational levels in H$_2$CS due to collisions with He atom, we have calculated cross sections under the CS approximation. In the MOLSCAT, there is provision to input more than one values of total energies. Here, for example, we are interested in the cross sections for total energy 11 cm$^{-1}$. The calculations have been done for the single energy 11 cm$^{-1}$ and for eight combinations, having energies (11, 12), (12, 11), (10, 11), (11, 10), (11, 12, 13), (9, 10, 11), (10, 11, 12), (9, 10, 11, 12, 13) cm$^{-1}$. We have found that the cross sections for 11 cm$^{-1}$, in general, differ from one another in all the 9 calculations. The reason for the difference in the results appears that the MOLSCAT uses the intermediate data of calculations for one energy, in the calculations for other energies. Under such circumstances, the possible suggestion can be to run the MOLSCAT for a single energy at a time.

astro-ph.IM↗
arXiv2014

Learning a hyperplane regressor by minimizing an exact bound on the VC dimension

The capacity of a learning machine is measured by its Vapnik-Chervonenkis dimension, and learning machines with a low VC dimension generalize better. It is well known that the VC dimension of SVMs can be very large or unbounded, even though they generally yield state-of-the-art learning performance. In this paper, we show how to learn a hyperplane regressor by minimizing an exact, or \boldmath{$Θ$} bound on its VC dimension. The proposed approach, termed as the Minimal Complexity Machine (MCM) Regressor, involves solving a simple linear programming problem. Experimental results show, that on a number of benchmark datasets, the proposed approach yields regressors with error rates much less than those obtained with conventional SVM regresssors, while often using fewer support vectors. On some benchmark datasets, the number of support vectors is less than one tenth the number used by SVMs, indicating that the MCM does indeed learn simpler representations.

cs.LG↗
arXiv2010

Linear magnetosonic waves in solar wind flow tubes

Nakariakov et al. (1996) investigated the linear magnetosonic waves trapped within solar wind flow tubes, where they accounted for a slab having boundaries at $x = \pm d$ and extended up to infinity in the $y$ and $z$ directions. Srivastava and Dwivedi (2006) claimed to extend that work by considering a two-dimensional slab. We find that the work of Srivastava and Dwivedi (2006) is not for a two-dimensional slab and has a number of discrepancies. Further, their results for body waves are not reliable.

astro-ph.SR↗
arXiv2009

Response to the comments of Dwivedi and Srivastava on the propagation and dissipation of Alfven waves in coronal holes

Chandra [1] made an attempt to show that the work of Dwivedi and Srivastava [2] (hereinafter DS) can be investigated even analytically and their results are erroneous. Dwivedi and Srivastava [3] picked up some values of Chandra [1] and tried to show that they are not physically acceptable. Some results of Chandra [1] are not physically acceptable, as these are the outcome of the wrong approach of DS. However, the results are numerically correct whereas the results of DS are numerically wrong.

astro-ph.SR↗
arXiv2009

Comment on "Dispersion relation for MHD waves in homogeneous plasma"

Pandey & Dwivedi (2007) again tried to claim that the dispersion relation for the given set of equations must be a sixth degree polynomial. Through a series of papers, they are unnecessarily creating confusion. In the present communication, we have shown how Pandey & Dwivedi (2007) are introducing an additional root, which is insignificant. Moreover, five roots of both the polynomials are common and they are sufficient for the discussion of propagation of slow-mode and fast-mode waves.

astro-ph.SR↗