Support vector regression (SVR) combined with particle swarm optimization (PSO) for its parameter optimization, was proposed to establish a model to predict the thermal conductivity of polymer-based composites under different mass fractions of fillers (mass fraction of polyethylene (PE) and mass fraction of polystyrene (PS)). The prediction performance of SVR was compared with those of other two theoretical models of spherical packing and flake packing. The result demonstrated that the estimated errors by leave-one-out cross validation (LOOCV) test of SVR models, such as mean absolute error (MAE) and mean absolute percentage error (MAPE), all are smaller than those achieved by the two theoretical models via applying identical samples. It is revealed that the generalization ability of SVR model is superior to those of the two theoretical models. This study suggests that SVR can be used as a powerful approach to foresee the thermal property of polymer-based composites under different mass fractions of polyethylene and polystyrene fillers.

According to the experimental dataset on the bending strength of AlON-TiN composite synthesized by hot pressing sintering approach under different processing parameters, i.e., mass fraction of TiN, sintering temperature and soaking time, the support vector regression （SVR） approach combined with particle swarm optimization for its parameter optimization, is proposed to simulate the relationship between the bending strength and hot pressing sintering synthesis parameters of AlON-TiN composites. The optimization of process parameters and the multi-factor analysis are also carried out. The prediction result demonstrates that the estimation error of the SVR model is less than that of the artificial neural network（ANN） model under the identical training and test samples and reveales that the generalization ability of SVR model surpasses that achieved by the ANN model. The optimal synthesis parameters are obtained numerically under TiN content 13.5%, sintering temperature 1863.5 ℃ and soaking time 5.8 h. The maximum bending strength is estimated to be 555.452 MPa while the AlON-TiN composite is synthesized at the optimal synthesis parameters. These results suggest that SVR can provide an important theoretical and practical guidance to the research and development of AlON-TiN composite possessing ideal bending strength.

This work presents a novel method of reducing on-resistance (R on) in vertical double diffusion metal oxide semiconductor (VDMOS) devices using external stress. By electrodepositing a layer of nickel (Ni) film onto the surface of drain pad, intrinsic film stress is applied on the VDMOS chip. The experiment results show that R on was reduced by 1.7, 2.4, 4.5, 8.3 and 10.5% when the film stresses were 39.2, 49.3, 61.7, 71.6 and 80.3 MPa, respectively. Moreover, thermal stress between Ni film and VDMOS chip can further reduce R on when the temperature increases. Compared with conventional methods for R on reduction, this method is independent of fabrication process of VDMOS devices and capable of controlling accurately the applied stress.