China Industry Of Foreign Anti-dumping Neural Network Of Early Warning Systems

With the development of the world's economy integration and international trade, trade friction is becoming an issue which is valuable to be paid attention to. In terms of China, although participating in international trade actively helps the development of China's economy and brings many advantages, the trade frictions against China caused by it are very frequent. Among them, anti-dumping is one of the most model trade frictions. Anti-dumping petitions against China which happens frequently becomes the main obstacle to Chinese export trade, and so early-warning system aimed at export commodities must be constructed immediately.The researches on the early-warning system of anti-dumping at home and abroad have gotten some achievements, but most of them focus on the early-warning system of anti-dumping against import commodities. At present, although many researchers studies on the issue about anti-dumping against China initiated by foreign countries and some of them refer to constructing early-warning system aimed at export commodities, the directing and the early-warning technology are not strong, and the availability is not good. Facing the rigorous reality that China is one of the most serious countries which suffered from anti-dumping in the world, it is necessary and valuable to construct an early-warning system aimed at export commodities. This thesis starts because of this.This thesis studies on the ANN early-warning system of anti-dumping against China's industry initiated by foreign countries, which is research on anti-dumping against China's industry initiated by foreign countries using ANN. According to this, we propose the early-warning index system of anti-dumping against China's industry initiated by foreign countries based on International Political Economy (IPE) by literature-reading and perfect it through expert inquiry. Then we construct the early-warning system of anti-dumping against China's industry initiated by foreign countries using ANN through the process of 'warning condition index-warning omen index-prediction of warning omen index by ANN (rolling prediction) -index significance-warning boundary-forecast of warning condition', and add the 'early-warning to measure' mechanism to improve the function of the ANN early-warning system of anti-dumping against China's industry initiated by foreign countries.Then we take anti-dumping against China's chemical industry initiated by USA as an example to demonstrate the ANN early-warning system of anti-dumping against China's industry initiated by foreign countries.Through the forecast effect test, we know that the ANN early-warning system of anti-dumping against China's industry initiated by foreign countries constructed in this thesis is effective. By it we can get the tendency of warning condition and the prediction of warning condition directly. And with the 'early-warning measure', the function of early-warning can be realized. It offers reference for constructing the early-warning system of anti-dumping.