关键词: 卷烟销售, 最小二乘支持向量机, 混合核函数, 布谷鸟搜索算法

Abstract: In order improve the forecast accuracy of cigarette sales, this paper proposes a cigarette sales forecasting model based on hybrid kernels Least Square Support Vector Machine（LSSVM） which parameters are optimized by Modified Cuckoo Search（MCS） algorithm. Firstly, the data of cigarette sales are collected to build the LSSVM sample, then the hybrid kernel function LSSVM is used to train samples, and the improved cuckoo algorithm is used to optimize the parameters of hybrid kernel function, finally, cigarette sales forecast model is established and the performance is tested by some cigarette sales data. The results show that, compared with other models, the proposed model has good modeling effect and higher forecasting accuracy.

Key words: cigarette sales, Least Square Support Vector Machine（LSSVM）, hybrid kernels, Cuckoo Search（CS） algorithm