Abstract
Abstract Inflation forecasting is one of the most indexes in economic situation. There are a lot of factors that affect inflation, such as the interest rate, exchange rate and the oil price and etc. When adding all of the variables, that will make the degree of freedom problem and interaction problem between the variables. This paper uses the principle component analysis to overcome the problem. And this paper uses Autoregressive model, Phillips curve model, and ordinary least square to compare with principle component analysis. Most of the principle component analysis make more accurate to forecast in inflation.