Finding an Effective Set of ATL Spots Learned from Questionnaire Data using a Sequential Minimal Optimization

• 발행기관: 한국자료분석학회
• 저자명: 김진상(Jin-Sang Kim) 신양규(Yang-Gyu Shin)
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.10 No.2, 613~623쪽, 전체 11쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2008.04.30

영문 초록

We apply a sequential minimal optimization method to business questionnaire data for finding the most cost effective set of spots in the above-the-line(ATL) marketing. The resulting set of important spots can be used in many ATL marketing activities to promote companies brand/image, specific products and services. Our method has been tested to various kinds of questionnaire data about cellular phones, TVs, supermarkets and auto- mobiles for its generality. One of the main findings of this paper is supervised learning can solve an optimal regression problem efficiently by using a sequential minimal optimization method. The other interesting result is a data filter to make a training data set s sheer size. Finally, not the least, our study is a real application that is used in decision-making system at a leading company in Korea.

목차

1. Introduction

2. Regression with Support Vector Machines

3. Methodology

4. Conclusions

References

키워드

Questionnaire data Sequential minimal optimization ATL marketing Regression problem