Feed-Forward Support Vector Machine Without Multipliers

In this letter, we propose a coordinate rotation digital computer (CORDIC)-like algorithm for computing the feed-forward phase of a support vector machine (SVM) in fixed-point arithmetic, using only shift and add operations and avoiding resource-consuming multiplications. This result is obtained thanks to a hardware-friendly kernel, which greatly simplifies the SVM feed-forward phase computation and, at the same time, maintains good classification performance respect to the conventional Gaussian kernel

[1]  Fred J. Taylor,et al.  A Fast Modified CORDIC—Implementation of Radial Basis Neural Networks , 1998, J. VLSI Signal Process..

[2]  Henk J. Sips,et al.  On-line CORDIC algorithms , 1989, Proceedings of 9th Symposium on Computer Arithmetic.

[3]  K.S.J. Pister,et al.  An ultra-low energy microcontroller for Smart Dust wireless sensor networks , 2004, 2004 IEEE International Solid-State Circuits Conference (IEEE Cat. No.04CH37519).

[4]  Yu Hen Hu,et al.  The quantization effects of the CORDIC algorithm , 1992, IEEE Trans. Signal Process..

[5]  Gunnar Rätsch,et al.  Soft Margins for AdaBoost , 2001, Machine Learning.

[6]  Davide Anguita,et al.  Learning algorithm for nonlinear support vector machines suited for digital VLSI , 1999 .

[7]  Jack E. Volder.  The CORDIC Trigonometric Computing Technique , 1959, IRE Trans. Electron. Comput..

[8]  D. Anguita,et al.  The effect of quantization on support vector machines with Gaussian kernel , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[9]  Bin Yu,et al.  Model Selection and the Principle of Minimum Description Length , 2001 .

[10]  Ralf Herbrich,et al.  Learning Kernel Classifiers , 2001 .

[11]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[12]  Behrooz Parhami,et al.  Computer arithmetic - algorithms and hardware designs , 1999 .

[13]  Ray Andraka,et al.  A survey of CORDIC algorithms for FPGA based computers , 1998, FPGA '98.

[14]  Deborah Estrin,et al.  Guest Editors' Introduction: Overview of Sensor Networks , 2004, Computer.

[15]  Davide Anguita,et al.  A digital architecture for support vector machines: theory, algorithm, and FPGA implementation , 2003, IEEE Trans. Neural Networks.

[16]  W. H. Specker.  A Class of Algorithms for Ln x, Exp x, Sin x, Cos x, Tan-1 x, and Cot-1 x , 1965, IEEE Trans. Electron. Comput..

[17]  Marc G. Genton,et al.  Classes of Kernels for Machine Learning: A Statistics Perspective , 2002, J. Mach. Learn. Res..

[18]  Mani Srivastava,et al.  Overview of sensor networks , 2004 .

[19]  S. Walther.  A unified algorithm for elementary functions , 1899 .

[20]  Manfred K. Warmuth,et al.  Relating Data Compression and Learnability , 2003 .

[21]  An-Yeu Wu,et al.  A unified view for vector rotational CORDIC algorithms and architectures based on angle quantization approach , 2002 .

[22]  Francesco Piazza,et al.  Fast neural networks without multipliers , 1993, IEEE Trans. Neural Networks.

[23]  Alexander J. Smola,et al.  Learning with kernels , 1998 .

[24]  William M. Pottenger,et al.  Finite precision analysis of support vector machine classification in logarithmic number systems , 2004 .

[25]  William M. Pottenger,et al.  Finite precision analysis of support vector machine classification in logarithmic number systems , 2004, Euromicro Symposium on Digital System Design, 2004. DSD 2004..

[26]  R. A. Smith.  Minimal Three-Variable NOR and NAND Logic Circuits , 1965, IEEE Trans. Electron. Comput..

[27]  J. Rissanen.  Stochastic Complexity in Statistical Inquiry Theory , 1989 .

[28]  Davide Anguita,et al.  Theoretical and Practical Model Selection Methods for Support Vector Classifiers , 2004 .

[29]  Tomás Lang,et al.  High-throughput CORDIC-based geometry operations for 3D computer graphics , 2005, IEEE Transactions on Computers.