Traditional classification algorithms perform not very well on imbalanced data sets and small sample size. To deal with the problem, a novel method is proposed to change the class distribution through adding virtual samples, which are generated by the windowed regression over-sampling (WRO) method. The proposed method WRO not only reflects the additive effects but also reflects the multiplicative effect between samples. A comparative study between the proposed method and other over-sampling methods such as synthetic minority over-sampling technique (SMOTE) and borderline over-sampling (BOS) on UCI datasets and Fourier transform infrared spectroscopy (FTIR) data set is provided. Experimental results show that the WRO method can achieve better performance than other methods.

No abstract

This work investigates the microstructure and inclusions of a compacted graphite iron (CGI) alloyed by Ce and La rare earth (RE) elements. In our study, alloying elemental distribution and solute segregation were characterized by methods of secondary ion mass spectrometry (SIMS) and a three-dimensional atom probe (3DAP) with high sensitivity and spatial resolution. RE sulfide, MgS, carbide, and composite inclusions formed during solidification and provided heterogeneous nucleation cores for the nucleation of the graphite. Significant solute clustering in the matrix, coupled with the segregation of solute to grain boundaries, was observed. C, Mn, Cr, and V were soluted in cementite and promoted the precipitation of cementite, while Si was found to be soluted in ferrite. Cu is usually distributed uniformly in ferrite, but some Cu-rich atom clusters were observed to segregate towards the interface between the ferrite and cementite, stabilizing the pearlite. In addition, P, as a segregation element, was enriched along the boundaries continuously. The RE elements participated in the formation of inclusions, consuming harmful elements such as As and P, and also promoted the heterogeneous nucleation of the graphite and segregated, in the form of solute atoms, at its interfaces.

No abstract

The impact behaviors of N‐bearing QN1803 and 304 stainless steels are studied using instrumented Charpy impact test in the range of −192–100 °C. The fracture and deformation substructures are systematically investigated by scanning electron microscopy and transmission electron microscopy. The results show that QN1803 presents ductile–brittle transition temperature (DBTT) and 304 shows ductile fracture. With increasing temperature, the stacking fault energy (SFE) of QN1803 and 304 increases, but the former is always lower than the latter. QN1803 presents low impact toughness with low‐density dislocation at −192 °C. At high temperature, QN1803 and 304 show good impact toughness and the deformation substructures are dominated by high‐density dislocations and deformation twinning.