人力資源分配管理之人工式世界電腦模擬硏究

摘要

人力資源的適度分配，將增加人們對環境的滿意度。傳統的研究著重在總體性的分析（也就是純粹是以人口數量來考量），而個別遷移者的行爲往往被忽略。這篇論文的主要目的,是擷取實際生活的「人」、「地」、「事」、「物」，建立一個人工世界的電腦模擬系統，以探討不同類型的「人」，在不同環境的「地」區，所發生的「事物」（遷移行爲）。實驗結果顯示，改善遷移的阻礙因素，可以提髙人們對環境的滿意度，但也導致人們產生聚集化。適度的反應人口增加所帶來的壓力，可以減少人口聚集化，但過度的強調這壓力將再度促使人口聚集化。最後，我們實驗結果說明人口遷移情形及人們的滿意度與環境大小及型態有相當大關係。

Abstract

Appropriate assignment of human resource would undoubtedly increase human satisfaction. Traditional approaches emphasize on the study in an aggregate manner (i.e., in term of the quantity of migration), whereas individual behaviors are totally neglected. The objective of this paper was to construct a computer simulation model that abstracts people, environments, and behaviors of the real world, and that investigates the behaviors of versatile peoples under different types of environments. Experimental results demonstrated that improving the obstacles of migration can increase humans' fitness, but will result in the clustering problem (i.e., people tend to gather in selective areas). The problem can be partially resolved by putting pressure on the quantity of people in each environment. However, overemphasis of the pressure would bring up the clustering problem again. Finally. we showed the distribution and the fitness of people are closely related to the sizes and types of environments.

關鍵詞:

• 人口遷移
• 電腦模擬
• 學習演算法
• 人口分佈

Key Words:

• human migration
• computer simulation
• evolutionary learning algorithm
• human distribution