Efficient kernel functions for support vector machine regression model for analog circuits’ performance evaluation

Abstract

Support vector machines (SVMs) have been widely used for creating fast and efficient performance macro-models for quickly predicting the performance parameters of analog circuits. These models have proved to be not only effective and fast but accurate also while predicting the performance. A kernel function is an integral part of SVM to obtain an optimized and accurate model. There is no formal way to decide, which kernel function is suited to a class of regression problem. While most commonly used kernels are radial basis function, polynomial, spline, multilayer perceptron; we have explored many other un-conventional kernel functions and report their efficacy and computational efficiency in this paper. These kernel functions are used with SVM regression models and these macromodels are tested on different analog circuits to check for their robustness and performance. We have used HSPICE for generating the set of learning data. Least Square SVM toolbox along with MATLAB was used for regression. The models which contained modified compositions of kernels were found to be more accurate and thus have lower root mean square error than those containing standard kernels. We have used different CMOS circuits varying in size and complexity as test vehicles-two-stage op amp, cascode op amp, comparator, differential op amp and voltage controlled oscillator.

Breast Cancer Detection using Support Vector Machine

“…However, the regional skewness and fractal dimensions had no association with breast cancer after adjusting for other risk factors and overall breast density. One feature, lacunarity, remained significant [12].…”

“…Methods such as hierarchical normalized cut, colour gradient active contour, colour texture cell Gaussian mixture model based segmentation and object graph approach were used for segmentation. Texture classification using fractal textures, classification using nonlinear quantization and support vector machine were used for classification approach [4], [7]- [11].Texture has been studied as a breast cancer risk factor independent of average breast density [12]- [17], but the results have not been adequately adjusted for breast density and other risk factors. For example, a negative significant correlation between regional skewness, fractal dimension, results cancer risk [13].…”

“…The resultant algorithm is similar with the exception that each dot product is substituted by a nonlinear kernel function. This permits the algorithm to adjust the maximum-margin hyper plane in an altered feature space [12].…”

An Empirical Analysis of Imbalanced Data Classification

“…TP and FP represent respectively the true and false positive class, while TN and FN stand respectively for the true and false negative class (Tang & Zhang, 2009). In Table 2, we provide the different performance metrics that are employed in the experimental task of data classification (Veropoulos, et al1999;Tang Zhang, 2009;Boolchandani Sahula, 2011). …”

“…A well-known (binary) classifier is the Support Vector Machine (SVM), which was initially introduced by Vapnik (Vapnik, 1998). SVM, a relatively new machine learning method, became very successful due to its strong theory and excellent performance in data classification (Haibo Garcia, 2009;Boolchandani Sahula, 2011). Additionally, SVM has shown remarkable success in various domains, such as pattern recognition and text classification.…”