Multivariate Adaptive Regression Splines in Standard Cell Characterization for Nanometer Technology in Semiconductor

Abstract

Multivariate adaptive regression splines (MARSP) is a nonparametric regression method. It is an adaptive procedure which does not have any predetermined regression model. With that said, the model structure of MARSP is constructed dynamically and adaptively according to the information derived from the data. Because of its ability to capture essential nonlinearities and interactions, MARSP is considered as a great fit for high-dimension problems. This chapter gives an application of MARSP in semiconductor field, more specifically, in standard cell characterization. The objective of standard cell characterization is to create a set of high-quality models of a standard cell library that accurately and efficiently capture cell behaviors. In this chapter, the MARSP method is employed to characterize the gate delay as a function of many parameters including process-voltagetemperature parameters. Due to its ability of capturing essential nonlinearities and interactions, MARSP method helps to achieve significant accuracy improvement.

Determination of Affecting Factors on Profitability of Banking Sector: An Examination upon the Turkish Banking Sector for the Period of 2006-2018

“…There are no assumptions between dependent and independent variables in the MARS method. In founding the effects of independent variables on dependent variables, MARS uses interactions between variables, and the effects of these interactions on dependent variables (Goh et al, 2017;Liu, 2018).…”

Defining Influential Factors of Capital Adequacy Ratio: An Examination upon Turkish Banking Sector (2006/Q1-2019/Q1)

“…There are no assumptions among dependent and independent variables in the MARS method. In searching the effects of independent variables on dependent variables, MARS also uses interactions between variables, and the effects of these interactions on dependent variables (Goh et al, 2017;Liu, 2018).…”

New voltage and temperature scalable gate delay model applied to a 14nm technology

“…The results are compared to a classical stastical method and achives more than 4% average error for all cells in the library. While this result is interesting, it seems quite high especially that tests on full circuit haven't been done [15]. Other works proposed other methodologies of standard cells characterization but they still require a high amount of running time [16][17][18].…”