• 沒有找到結果。

來台旅客消費結構分析

N/A
N/A
Protected

Academic year: 2021

Share "來台旅客消費結構分析"

Copied!
4
0
0

加載中.... (立即查看全文)

全文

(1)

作者/Author:

頁數/Page:

出版日期/Publication Date:

To cite this Article, please include the DOI name in your reference data.

引用本篇文獻時,請提供DOI資訊,並透過DOI永久網址取得最正確的書目資訊。

請使用本篇文獻DOI永久網址進行連結:

To link to this Article:

DOI是數位物件識別碼(Digital Object Identifier, DOI)的簡稱, 是這篇文章在網路上的唯一識別碼,

用於永久連結及引用該篇文章。 若想得知更多DOI使用資訊, 請參考 http://doi.airiti.com

http://doi.airiti.com

For more information, Please see:

來台旅客消費結構分析

The Analysis of Consumption Structure for Foreign Tourists in Taiwan

doi:10.6132/JCM.2007.4.2.03

商業現代化學刊, 4(2), 2007

Journal of Commercial Modernization, 4(2), 2007

鄒幼涵(You-Han Tsou);李仁棻(Jen-Fen Li);黃登源(Teng-Yuen Huang)

23-30

http://dx.doi.org/10.6132/JCM.2007.4.2.03

(2)

You-Han Ts凹, Jen-Fen Li, Teng-Yuen Huang / Joumal of Commercial Modernization, Vo1.4

,

No.2

,

(2帥7)23品

來台旅客消費結構分析

鄒幼涵

李仁菜*

黃登源** 輔仁大學裔學研究所 *高雄應用科技大學會計系 "輔仁大學應用統計研究所 摘要 交通部觀光局每年均舉辦來台旅客消費及動向調查以便了解旅客對我國觀光資源、環境之認知 程度、其來台期間旅遊之動向及消費行為,而估算我圍觀光外匯收入為調查主要目的之一。深度探 討影響旅客此次來台消費金額之變數,分析不同消費族群之結構特性,建構合適模型,可應用於調 查資料遺漏值之插補。本研究依據歷年觀光局「觀光統計年報」及 「來台旅客消費及動向調查」資 料,應用於顧客消費金額之結構分析中,建立適當的邏輯斯迴歸模型,過程中採用基因演算法 , 並 以傳統方法為比較基礎,獲得整體最適解,並進行模型評估及精筒。本文分析結果發現﹒年齡與過 夜日數是影響來台旅客消費金額高低的關鍵特質 。 瞭解消費金額高低之顧客的組成要素後,將可以 作為行銷策略接訂時之參考 。 關鍵字:順序邏輯斯迴歸 ,基因演算法,決策樹,自動互動檢視法,赤池資訊準則。 1.緒言 交通部觀光局每年均舉辦來台旅客消費及 動向調查以便了解旅客對我圍觀光資源、環境 之認知程度、其來台期間旅遊之動向及消費行 為等資訊。深度探討影響旅客此次來台消費金 額之變數,分析不同消費族群之結構特性,這 對我國發展觀光事業將有很大幫助。在「九十 三年來台旅客消費及動向調查」之抽樣設計 中,調查主要目的為估算我國觀光外匯收入, 將影響來台人次的五個變數,居住地、來台目 的、性別、年齡和月份作為“抽樣控制特徵“。 由於五個變數的組合共有 3,240 種,基於調查 成本及可行性的考量,因此本調查依據歷年觀 光局「觀光統計年報」 及「來台旅客消費及動 向調查」資料,採用自動互動檢視法 (Automatic

Interaction Detector, AID) (Breiman et., (1998))

來控制產生互動的變數,推算影響來台人次之 抽樣母體結構,採用配額抽樣法,取樣時儘量 符合作業標準及樣本比例,經由分析後僅需控 制 225 種組合,透過這些計算得到五個變數組 合的樣本分配數,便消費金額差異越大之旅客 群抽取更多樣本數,反之越少,以達到提高抽 樣效率,減少成本之目的。 本文利用來台旅客觀光調查資料(黃登源 等, (2∞4)) ,應用於顧客消費金額之結構分析 中,建立適當的邏輯斯迴歸模型(Logistic Regression Models) ,邏輯斯迴歸分析是處理分 類判別問題很重要的方法,它可以將各類別之 結構顯示出來。近年來,由於計算工具之效率 提升,對於有順序之類別資料之計算,已經可 以克服更復雜的計算,對於參數估計及變數之 精簡σarsimonious輝有很大的幫助。 邏輯斯迴歸法傳統上使用數學演算法(黃 令、 d 、 ' ,

(3)

You-Han Ts凹, Jen-Fen Li, Teng-Yuen Huang I Joumal of Commercial Modernization

,

Vol.4

,

No.2

,

(2帥7)23品 參考書籍 1. 王濟川、郭志剛 , rLogistic 迴歸模型一 方法及應用 J 五南圖書出版股份有限公 司 , pp86-冊, 2∞4年9月 。 2. 黃登源 r 應用迴歸分析J 華泰書局, 1998年 。 3. 資登源等 r 民國九十三年來台旅客消費 及動向調查報告 J 交通部觀光局 , 2∞4 年6月 。 4. 蘇木春、 張孝德 r 機械學習 、類神經網 路、模糊系統以及基因演算法貝目 J 全華 科技圖書股份有限公司, 1999年 。 5. 6.

Agresti, A. Analysis of ordinal categorical

da徊, John Wiley & Sons, Inc. pp112-132.

(1984).

Ak位尬, H. Information theory and an extension of the maximum likelihood principle. In B. N. Pe虹ov & F. Csaki (吋),

7.

8.

9.

Second International Symposium on

Information Theory (pp271 月281). Budapest:

Akademiai Kiado. (1973).

Answer Tree 2.0 Users Guide. SPSS Inc.

(1998). ppl-65.

Breiman, L., Friedm徊, J. 旺, O1shen, R. A.

組d Ston巴 C. J. Classificationωd

regression 甘ees. Chapman and HalIlCRC,

BocaRato且, F1a. (1998).

Cox, D. R. and Snell, E. J. The Analysis of nd

Binary Data. (2 ed.). London: Chapman and Hall. (1989).

10. Dale到us, T. and Hodg郎, J. L., Jr.

Minimum Variance Stratification. Jo叮nal

of American Statistical Association. Vol. 54,

pp88-101. (1959).

(4)

You-H個 Tsou, Jen-Fen Li, Teng-Yuen Huang / Joumal ofCommercial Modemization,

Vol.4

,

No.2

,

(2個7)23品

The Analysis of Consumption Structure for Foreign Tourists

in Taiwan

*You-Han Tsou

,

**Jen-Fen

Li

,

***Teng-Yuen Huang

ABSTRACT

To help understand the motives, the recognition of environment,由e tendency of 回veling and 由e

consuming behaviors of foreign to叮ists in Taiw妞,也eM姐IS虹y of Communications of Taiwan Tourism Bureau conducts the investigations of foreign to叮ists' consumption and tendency annua11y. In addition, the estimation of foreign exchange revenue is a1so one of the main p凶poses. In recent ye訂s, T訟wan

Tourism Bureau has made every effort in providing 也e adequate tourist information and good sightseeing facilities wi血 successfu1 marketing strategies 伽t can enormously promote the Taiw妞's image in intemationa1 sightseeing. Based 翎0nT剖w組 T芯0叮叩ism Bu叮Ire翎渴u'池s

‘“‘t曲he Annua1 Report of the Investigation of Foreign Tourists' Consumption and Tendency", this study

ana1yzes the structure characteristics of 也e distinct consumer grou阱, bui1ds a model, and applies it to 由e

investigation where the interpolation of missing data va1ues may occur. Accordingly, through the

fr組lework of customers' consumption level, the model of logistic regression was bui1t, the method of genetic a1gorithm was applied, and when compared to 虹.aditiona1 me也od 血e optimal solution was ac凶eved, which inc1udes model eva1uation and sirnplification. 四le findings suggest that bo也 age and days of stay are most critica1 the factors influencing foreign touris毯, consumption level in terms of

money 由ey spent. It is noted 血at understanding of foreign touris尬, consumption level can help 血e

formulation of marketing s甘ategy for to山1sm or traveling comp缸茵的.

Keywords: Logistic Regression, Genetic A1gorithms, Consuming Behaviors

參考文獻

相關文件

Milk and cream, in powder, granule or other solid form, of a fat content, by weight, exceeding 1.5%, not containing added sugar or other sweetening matter.

(c) Draw the graph of as a function of and draw the secant lines whose slopes are the average velocities in part (a) and the tangent line whose slope is the instantaneous velocity

You are given the wavelength and total energy of a light pulse and asked to find the number of photons it

Reading Task 6: Genre Structure and Language Features. • Now let’s look at how language features (e.g. sentence patterns) are connected to the structure

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

incapable to extract any quantities from QCD, nor to tackle the most interesting physics, namely, the spontaneously chiral symmetry breaking and the color confinement.. 

Based on the reformulation, a semi-smooth Levenberg–Marquardt method was developed, and the superlinear (quadratic) rate of convergence was established under the strict