Forecast for the makeup blog users care products ad type preferences- The use of neural networks

• Main Authors: Ping-Chuan Lee, 李蘋娟
• Other Authors: 潘振雄
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/72260352389071785784

碩士 === 清雲科技大學 === 企業管理系暨經營管理研究所 === 97 === The internet advertising for website and the click rate is not high enough to compete with the way of traditional advertising. This study tries to use the advantages of neural system to handle non-linear and noise information. Moreover, it develop the neural system prediction model and use the lifestyle variables to predict the makeup blog users advertising preference. In the future, internet advertising can applied this model to offer the effectiveness advertising page to classified customers. The convenience sampling method used in this study and the survey placed in three different discussed blog. The total of six hundred surveys is returned. The blog user''s lifestyle factors are divided into twelve dimensions after analysis from return questionnaires; the makeup blog user’s advertising type of skin care products preference have nine dimensions. The dimensions of the various factors point as the neural modeling data. Also the development of model is expected to the basic user blog lifestyle scale on the skin care products for expected their preferred type of advertising. The results of the neurons hidden layer number set to eight, learning rules for Tanh or Gaussian has the minimum error rate. However, it has no significant results after increase the number of hidden layer neurons. This mode will enable to make the cooperation and benefit between the advertisers, the advertising industry and users from the three parties.