Ridge regression is known to be an effective algorithm for regression problems. However, the algorithm has some drawbacks with highly non-linear datasets. In this research, the hot deformation flow stress of 321 stainless steel was modeled using the kernel ridge regression algorithm. For modeling the flow stress in this research, the tensile test data for 321 stainless steel under temperatures of 700℃, 800℃, and 900℃ at strain rates of 0.0002/s, 0.002/s, and 0.02/s were used. To overcome the drawbacks of the traditional ridge regression algorithm, the algorithm was enhanced by a kernel-type function to handle the non-linear dataset. The predicted data by the kernel ridge regression was accurate. After that, the predicted values were studied in terms of their distribution. The kernel ridge regression algorithm was found to be accurate and stable in predicting the flow stress of hot deformation.

Abstract
1. 서 론
2. 릿지 회귀(Ridge regression)
3. 커널 릿지 회귀(Kernel ridge regression)
3. 비교분석
4. 결 론
References

• 송신형(Professor, Dept. of Smart Automobile, Soonchunhyang University) | Song Shin-Hyung Corresponding author