本文關鍵詞： 消費者行為 支持向量機 潛在購買力 保健品 仿真技術(shù)　出處：《安徽理工大學》2013年碩士論文　論文類型：學位論文

【摘要】：近年來,保健品行業(yè)和消費者行為均為學術(shù)研究的熱點。如何利用現(xiàn)有的技術(shù)和理論,實現(xiàn)數(shù)據(jù)仿真,構(gòu)建行為模型,挖掘潛在保健品消費群體和潛在購買力將成為保健品行業(yè)的重要發(fā)展方向。本文中的保健品包括食用的蜂制品、維生素等一般營養(yǎng)品,具有保健功能的保健藥品、藥膳和保健化妝品。保健品的消費情況不同于一般的生活日用品,而且其消費更易受更多的因素影響,需要探索一條適合保健品消費者行為的研究途徑。 文章首先簡單介紹了課題研究的背景、意義、目的和方法等,并論述了國內(nèi)外消費者行為的研究現(xiàn)狀,指出國內(nèi)研究起步晚、研究成果不多、研究方法有待創(chuàng)新。分析了我國保健品市場經(jīng)歷了起步階段、快速發(fā)展階段、蕭條發(fā)展階段和復蘇調(diào)整階段。在短短十幾年時間里,保健品已經(jīng)迅速發(fā)展成為一個獨特的、備受各界矚目的熱點產(chǎn)業(yè),成為中國工業(yè)經(jīng)濟新的增長點和國民經(jīng)濟的新興行業(yè)。盡管我國保健品行業(yè)發(fā)展?jié)摿Υ�、產(chǎn)品質(zhì)量和科技含量不斷提高、具備一定的規(guī)�；�,但仍面臨著一些挑戰(zhàn)。保健品未來的發(fā)展將呈現(xiàn)消費者群體擴大化、營銷模式專營化、產(chǎn)品材料多樣化、產(chǎn)品類型多元化,市場監(jiān)控規(guī)范化的趨勢。文中研究了支持向量機和BP神經(jīng)網(wǎng)絡算法的基本原理和學習訓練過程,分析了兩種算法的優(yōu)缺點。參考國內(nèi)外研究成果和問卷調(diào)查的基本原則,設計了適合保健品消費者行為研究的問卷,并針對有購買保健品意向的網(wǎng)民消費者發(fā)布問卷,收集有效問卷。利用SPSS19.0對收回的問卷信息進行描述性分析和信度、效度分析,表明問卷具有一定的實用性。本文將處理后的影響保健品消費者行為的各種重要因素即性別、年齡、收入等,通過仿真環(huán)境,進行回歸預測,并對具體流程做了詳細的描述。實驗表明,支持向量機算法在保健品消費者的購買能力和實際購買的預測中,比神經(jīng)網(wǎng)絡算法具有更高的準確性。 最后得到以下結(jié)論：支持向量機算法在解決小樣本、高維模式和非線性問題上具有較好的逼近能力和泛化能力；影響保健品消費者行為的主要因素有收入、節(jié)省程度、職位類別和婚姻狀況等；有購買意向的保健品消費者平均的實際購買占購買能力的2/3,具有一定的可挖掘性,且購買能力越高的消費者,其可挖掘性越大。
[Abstract]:In recent years, health products industry and consumer behavior have been the hot topics of academic research. How to use existing technology and theory to realize data simulation and build behavior model, It will be an important development direction for the health products industry to tap the consumption groups and potential purchasing power of potential health products. The health products in this paper include bee products, vitamins and other general nutrients, health drugs with health care functions. The consumption of health products is different from that of general daily necessities, and its consumption is more easily affected by more factors. Therefore, it is necessary to explore a research approach suitable for the consumer behavior of health products. Firstly, the paper briefly introduces the background, significance, purpose and methods of the research, and discusses the current research situation of consumer behavior at home and abroad, pointing out that the domestic research started late, and the research results are few. The research methods need to be innovated. It is analyzed that the health care products market in our country has experienced the initial stage, the rapid development stage, the depression development stage and the recovery adjustment stage. In a short period of ten years, the health care products have developed rapidly into a unique, Hot industries, which have attracted much attention from all walks of life, have become a new growth point of China's industrial economy and a new industry of the national economy. Despite the great potential for development of the health products industry in China, the quality of products and the scientific and technological content have been continuously improved, with a certain scale. However, there are still some challenges. The future development of health products will present the expansion of consumer groups, the monopoly of marketing mode, the diversification of product materials, and the diversification of product types. This paper studies the basic principle and learning and training process of support vector machine and BP neural network algorithm, analyzes the advantages and disadvantages of the two algorithms. This paper designed a questionnaire suitable for the study of consumer behavior of health products, and issued a questionnaire for consumers with intention to purchase health care products, and collected valid questionnaires. SPSS19.0 was used to analyze the information collected from the questionnaires in terms of descriptive analysis, reliability and validity analysis. The results show that the questionnaire has a certain practicability. In this paper, all kinds of important factors, such as gender, age, income and so on, which affect the consumer behavior of health products after treatment, are predicted by the simulation environment. The experimental results show that the SVM algorithm is more accurate than the neural network algorithm in predicting the purchasing ability and actual purchase of health products consumers. Finally, the following conclusions are obtained: the SVM algorithm has better approximation and generalization ability in solving small samples, high dimensional patterns and nonlinear problems, and the main factors affecting the consumer behavior of health products are income, saving degree, etc. The average actual purchase of health care products with purchase intention accounts for 2 / 3 of the purchasing capacity, which has a certain degree of diggability, and the higher the purchasing power, the greater the excavability of consumers.
【學位授予單位】：安徽理工大學
【學位級別】：碩士
【學位授予年份】：2013
【分類號】：TP181;F719

【參考文獻】

相關期刊論文 前10條

1 左洪亮;對影響消費者購買行為的心理因素的研究[J];商業(yè)研究;2005年10期

2 潛鐵宇;李慧子;;論女性保健食品包裝設計定位的重要性[J];包裝工程;2010年16期

3 宋志國;宋麗娟;;基于消費者行為理論的房地產(chǎn)綠色營銷分析[J];中國管理信息化;2008年18期

4 龍曉楓;田志龍;;影響消費者行為的規(guī)范理性[J];湖南大學學報(社會科學版);2010年05期

5 陳慧卿;;消費者購買行為分析[J];商業(yè)經(jīng)濟;2006年01期

6 朱存梅,王軍,計艷;消費者人格特質(zhì)及消費者行為研究[J];經(jīng)濟問題探索;2004年12期

7 周生超;居玲華;;基于市場細分的旅游競爭力分析[J];經(jīng)濟研究導刊;2008年16期

8 劉良林;王全鳳;林煌斌;;BP神經(jīng)網(wǎng)絡參數(shù)設定及應用[J];基建優(yōu)化;2007年02期

9 趙青松;李興兵;唐小松;;基于支持向量機的煙葉感官品質(zhì)評價[J];計算機工程與應用;2007年10期

10 楊新斌;黃曉娟;;基于支持向量機的股票價格預測研究[J];計算機仿真;2010年09期

相關博士學位論文 前1條

1 常甜甜;支持向量機學習算法若干問題的研究[D];西安電子科技大學;2010年

，

本文編號：1537389