Research of Household Savings Prediction Based on SVM and K-CV
In order to improve the prediction accuracy of household savings, this paper puts forward a household savings model based on support vector machine (SVM) and K-fold cross-validation (K-CV). First, it builds the sample of the historical data on household savings. Second, it pretreats the sample data, including normalized and principal component analysis (PCA) dimensionality reduction process. Third, it uses K-CV to select the optimal parameters. Fourth, it uses the best parameters to train the training set data. Finally, it predicts and analyzes the predictive set data and establishes the prediction model. The experimental results show that the prediction method of household savings has a higher accuracy than traditional method and has high generalization ability.
KeywordsSupport vector machine K-fold cross-validation Household savings
- 1.Shi F, Wang X-C, Yu L, Li Y (2010) 30 cases analysis of MATLAB neural network, vol 245, Beijing University Press, Beijing, pp 246–247Google Scholar
- 2.Chang C-C, Lin C-J (2010) LibSVM: a library for support vector machines, vol 63. http://www.csie.ntu.edu.tw/~cjlin/libsvm/. pp 1346–1347, 22 March 2010
- 3.Zheng X, Qian F (2006) Support vector machine with Gauss Kernel classified method and model selection. Comput Eng Appl 01(3):77–79Google Scholar
- 4.Zhang Y, Yin T-F (2011) Study on tax forecasting base on principal component analysis and support vector machine. Comput Simul 28(9):358–360Google Scholar
- 5.Dai H, Wang J, Chen Y-T (2011) The application of face recognition used fuzzy support vector machine. Comput Eng Appl 04(6):142–148Google Scholar