臺灣博碩士論文加值系統

隨著科技發展,人們逐漸進入AI時代,透過機器學習,各種工作漸漸地能夠自動化完成,而原本在該工作崗位的人(下稱舊有工作者)將面臨被取代的命運。目前的機器學習多為以使用資料來訓練為主流,但一般人不易取得大量資料,或是還未存在公開資料,並且舊有工作者並非對自動化領域熟悉,難以自行建立自動化機制;或者,當機器學習領域之專家或學者要與舊有工作者合作時,舊有工作者可能會考量到自己未來的工作被取代,使得自動化不易發展,因此,如果有一套方法能夠讓舊有工作者自行建立自動化機制,不失為一種考量。再者,考量到歷來已有多種優化方法藉以尋求最佳解,因此本研究將不著重於尋求最佳解,而是著重在使用者可自行建立自動化代理人。

As the development of technology, people step into AI era. Several tasks can be done automatically. People on his or her jobs are facing replacement. Nowadays, most machine learning mechanisms are data-driven. However, it's hard to obtain a large scale or a non-public dataset for an average person, and he or she may not be familiar with automation mechanism. Thus, having a method to let an average person build an automation mechanism might be a consideration. As there are several optimization techniques, this research will not focus on finding optimal solutions but letting users be able to build an automation mechanism by themselves.

中文摘要: i
英文摘要: ii
誌謝: iii
目錄: iv
圖目錄: vi
表目錄: vii
第一章 導論: 1
1.1 研究背景與動機: 1
1.2 研究目的: 1
1.3 研究範圍: 1
1.4 研究步驟: 2
1.5 論文架構: 2
第二章 相關背景知識: 2
2.1 堆積(heap): 2
2.2 廣度優先搜尋(Beadth-First Search, BFS): 3
2.3 A*: 3
2.4 Classical planning: 4
2.5 15-puzzle: 4
第三章 系統規劃與設計: 5
3.1 需求: 5
3.2 規劃與設計: 5
第四章 系統實作: 7
4.1 概觀: 7
4.2 使用者介面語法暨高層觀點描述之定義: 7
4.3 實作: 9
4.4 實驗暨應用範例: 11
4.4.1 將本系統應用於指揮mBot的行動: 11
4.4.2 將本系統應用於解15-puzzle: 16
第五章 結論與未來方向: 21
5.1 結論: 21
5.2 未來方向: 21
參考文獻: 22
附錄一 解 15-puzzle 用的高層觀點描述: 23

1. Sebastian Ruder, "An overview of gradient descent optimization algorithms" pp. 4-5, v2, 15 Jun 2017. Available: https://arxiv.org/abs/1609.04747
2. J. Kennedy, R. Eberhart, "Particle swarm optimization", IEEE International Conference on Neural Networks, pp. 1942–1948, 1995.
3. J. W. J. Williams, "Algorithm 232 - Heapsort", Communications of the ACM, vol. 7 no. 6, pp. 347–348, Jun, 1964.
4. Edward F. Moore, "The Shortest Path through a Maze", Proceedings of the International Symposium on the Theory of Switch, Part II, pp. 285-292, 1957.
5. Peter E. Hart, et al., "A Formal Basis for the Heuristic Determination of Minimum Cost Paths", IEEE Transactions on Systems Science, Cybernetics, vol. 4, no. 2, pp. 100-107, Jul, 1968.
6. Malik Ghallab, Dana Nau, Paolo Traverso, Automated Planning and Acting, Cambridge University Press, Jul, 2016.
7. Stuart J. Russell, Peter Norvig, Artificial Intelligence A Modern Approach Third Edition, pp. 377, Pearson Education, Dec, 2009.
8. https://en.wikipedia.org/wiki/15_puzzle
9. Mark Ryan, "Solvability of the Tiles Game", last modified: 26 November 2004. https://www.cs.bham.ac.uk/~mdr/teaching/modules04/java2/TilesSolvability.html