基于模糊神经网络的移动产品投放渠道评估研究

发布时间：2018-04-14 15:38

本文选题：渠道评估 + 模糊评价　；参考：《首都经济贸易大学》2017年硕士论文

【摘要】：随着移动互联网技术的发展和人们生活方式、生活理念的改变,在这个移动互联网时代的下,人们的生活状态和工作方式正在由先前的PC端大范围地、逐渐的转移到了移动端。针对企业在选择移动产品投放渠道的方式方法较为欠缺的问题,提出了基于模糊神经网络的移动产品投放渠道的评估方案。传统的投放渠道选择主要是基于价格和经验较为单一的指标来考量一个渠道是否具有投放价值,而由于传统的价格指标和经验的评判两个指标的不完善,导致其很难科学全面的评价一个渠道的优劣整体指标。本文结合传统的渠道选择评判方法,将评判指标从价格和专家经验扩展到更为完善和科学的层面,进一步提高了企业评判和选择投放渠道的科学性和全面性。针对神经网络仅凭数据进行处理和预测的盲目和依赖性以及专家经验评估的主观性等问题,提出结合专家经验和人工神经网络相结合的思路。这既可以保证了专家经验的优势发挥,同时也保证了数据验证的科学性。本文以某互联网公司为例,将企业实际的需求提炼为研究模型。从移动产品投放渠道的视角,根据历史的各个产品投放渠道的下载、稳定性、成本数据指标构建渠道投放模糊综合评价,结合专家的评分数据指标,将客观实际的数据指标和专家的经验评分有机的结合,为移动产品的广告投放提供更加科学合理的解释。运用模糊神经网络技术进行数据的输入、训练、仿真进而通过MATLAB得出具体的评估模型,为相关企业投放移动产品广告提供建议和参考依据指标。本文首次将模糊神经网络这一技术与移动互联网产品渠道投放这一细分领域的评估结合在一起,并根据传统的价格成本等理论进行优化和改造以适应该细分领域的渠道评估。在技术实现上,主要应用模糊神经网络原理,以该互联网公司部分移动产品投放数据为样本数据,通过MATLAB函数工具箱建立了ANFIS模型,经过仿真得到评价结果。结果表明,该方法成功的将模糊神经网络技术运用到互联网移动产品的投放效果评估和渠道评估预测。
[Abstract]:With the development of mobile Internet technology and the change of people's life style and concept of life, people's living state and working style are gradually transferred from PC to mobile end in this era of mobile Internet.In order to solve the problem that enterprises are lacking in the ways and methods of choosing mobile product delivery channels, an evaluation scheme of mobile product delivery channels based on fuzzy neural network (FNN) is proposed.The traditional channel selection is mainly based on the price and experience of a single index to consider whether a channel has the value of delivery, but because of the traditional price index and the evaluation of the two indicators of experience is not perfect.As a result, it is difficult to evaluate the overall quality of a channel in a scientific and comprehensive way.Combined with the traditional channel selection evaluation method, this paper extends the evaluation index from price and expert experience to a more perfect and scientific level, and further improves the scientific and comprehensive evaluation of enterprise evaluation and choice of investment channel.In view of the blind and dependence of neural network processing and prediction based on data and the subjectivity of expert experience evaluation, the idea of combining expert experience with artificial neural network is put forward.This not only ensures the advantage of expert experience, but also ensures the scientific nature of data verification.Taking an Internet company as an example, this paper abstracts the actual demand of an enterprise into a research model.From the perspective of mobile product delivery channel, according to the download, stability and cost data index of each historical product delivery channel, the fuzzy comprehensive evaluation of channel delivery is constructed, and the expert scoring data index is combined.The combination of objective and practical data index and expert's experience score provides a more scientific and reasonable explanation for advertising of mobile products.The fuzzy neural network technology is used to input, train and simulate the data, and then the concrete evaluation model is obtained by MATLAB, which provides suggestions and reference indexes for the advertising of mobile products by related enterprises.In this paper, the technology of fuzzy neural network is first combined with the evaluation of channel delivery of mobile Internet products, and is optimized and modified according to the traditional pricing and cost theory to adapt to the channel evaluation in this subdivision field.In technical realization, the principle of fuzzy neural network is mainly applied to build the ANFIS model through the MATLAB function toolbox, and the evaluation results are obtained by simulation based on some mobile product launch data of the Internet company as sample data.The results show that the fuzzy neural network technology is successfully applied to the evaluation of the effect of Internet mobile products and channel evaluation and prediction.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F713.8;TP183

【参考文献】