In this article, we fabricated amorphous InSnO thin film transistors (TFTs) with exceedingly high mobility and low thermal budget. The device is annealed only at a low temperature of 150 °C, a field-effect mobility ([Formula: see text]) of 70.53 cm2/V s, a subthreshold swing of 0.25 V/decade, an on/off current ratio over 108, and a reasonable threshold voltage shift under negative bias stress. The influence of thermal annealing on amorphous InSnO TFTs was investigated by systematically analyzing the crystallization, surface morphology, internal chemical state, and energy band relationship of the InSnO thin film. Amorphous InSnO films deposited at room temperature have a sparse and porous loose structure, which has carrier scattering caused by poor film quality, resulting in low mobility and few free carriers in the film. With the increase in the annealing temperature, the In and Sn metal cations are further oxidized, increasing the carrier concentration in the film and forming a dense M–O–M network when annealed at 150 °C. With the further increase in the annealing temperature, a large number of thermally excited free electrons make the device appear metal like conductivity. This paper expands the research on a high electron concentration InSnO material as the active layer and promotes the development of amorphous oxide semiconductors in high mobility and flexible TFTs.

No abstract

No abstract

With the change of people's consumption mode, credit consumption has gradually become a new consumption trend. Frequent loan defaults give default prediction more and more attention. This paper proposes a new comprehensive prediction method of loan default. This method combines convolutional neural network and LightGBM algorithm to establish a prediction model. Firstly, the excellent feature extraction ability of convolutional neural network is used to extract features from the original loan data and generate a new feature matrix. Secondly, the new feature matrix is used as input data, and the parameters of LightGBM algorithm are adjusted through grid search so as to build the LightGBM model. Finally, the LightGBM model is trained based on the new feature matrix, and the CNN-LightGBM loan default prediction model is obtained. To verify the effectiveness and superiority of our model, a series of experiments were conducted to compare the proposed prediction model with four classical models. The results show that CNN-LightGBM model is superior to other models in all evaluation indexes.