This paper aims to explore the effectiveness of an Importance–Performance Analysis approach to assess destination image. It focuses on the image of the UK for Chinese students studying in the country. This is because the destination image of a certain country for a homogenous group, such as Chinese students, has not been studied enough, and this specific market is central for the UK, especially its education and tourism sector. In total, 23 attributes were examined, and each was found in one of the five quadrants. Two attributes related to the environmental aspect of the UK were found in the Competitive Attributes quadrant. These attributes are key strengths of the UK in relation to Chinese students’ images of the country. Three attributes placed in the Keep Up Good Work quadrant are associated with British culture and heritage. This signifies that the UK should keep utilising these resources to improve Chinese students’ images of the nation. Fourteen attributes were identified in the Concentrate Here quadrant. These attributes include essential elements of tourism such as local cuisine and transportation. Their roles are vital in enhancing the image of the UK for Chinese students, but more efforts must be made to this end.

No abstract

PM2.5 is a kind of data with strong time-series characteristics, so the current PM2.5 prediction methods mostly choose RNN, LSTM and other sequence models for prediction, but because RNN, LSTM and other models use the same weight to calculate the data input at different times, which does not conform to the brain-like design, resulting in a low accuracy of PM2.5 concentration prediction. In view of the above problems, a PM2.5 prediction method based on attention mechanism (at-rnn and at-lstm) is proposed. This method firstly establishes the encoder-decoder model based on deep learning. In the Encoder stage, attention mechanism is added, and attention weight is allocated to the input with time series characteristics, and then Decoder analysis and prediction are carried out. Through experiments, the effects of RNN, LSTM, at-rnn and at-lstm on the prediction of PM2.5 concentration in hefei are compared. The results show that the accuracy of the prediction method based on attention model is better than other methods, indicating the effectiveness of the prediction method based on attention model in the prediction of pollutants.

Abstract. In recent years, with the slowdown of Chinese economic development and the demand for industrial transformation and upgrading, the real estate industry has also entered the stage of slow development and industry rectification. This paper tries to estimate the construction project cost index system by selecting the specific index as input variable and establishing the sample training set and constructing the RBF artificial neural network forecasting model under MATLAB environment. By comparing and analyzing with the actual data, this paper proves the practicability of the model.

No abstract