數據科學

數據科學（粵拼：sou3 geoi3 fo1 hok6 | 英文：data science）係一個跨學科科學領域，重點研究運用科學方法同各種演算法，由有結構或者冇結構嘅 data 嗰度搵出有用知識^[1]^[2]。

有數據科學家指，數據科學可以用以下一句嘢概括^[3]：用最勁嘅硬件、最勁嘅編程系統、同用最有效嘅演算法嚟解決問題。

諗頭

有人指，數據科學係科學嘅第四範式（fourth paradigm）：本嚟科學係以實證、理論同計算為重嘅，廿一世紀初就喺呢三樣之上加多個「數據」^[4]；數據科學會運用數學、統計學、機械學習、以及訊息科學等領域嘅技術，由手上嘅數據當中抽取有用嘅知識，並且攞嚟喺有經濟價值嘅地方嗰度使用（例：根據手上數據，呢柞變數之間成噉嘅關係，呢個預測現象嘅能力可以作乜商業用途？）^[5]。數據科學係伴隨廿一世紀資訊科技而嚟嘅－隨住互聯網發達，有大量數據喺網上流動（睇埋大數據），於是有唔少人開始想利用呢啲數據做有用嘅嘢，數據科學就隨住呢股趨勢而興起^[6]^[7]。

工序

做數據科學嘅工序如下^[8]：

諗清楚要解決嘅問題係乜。例：家陣有個數據科學家，俾人請佢幫手做市場研究，佢首先要搞清楚份研究想達到乜目的（搞清楚消費者鍾意同唔鍾意件產品啲乜，再按呢個資訊改良件產品）。
做一啲探索性質（exploratory）嘅分析，諗吓手上嘅數據解答到乜嘢同目的相關嘅問題。例：手上個數據庫有一大柞消費者嘅數據，知道每個消費者對件產品喺各方面嘅評價，計劃可以試吓做聚類分析將消費者分做若干類。
建立一啲數學模型，描述柞數據入面有啲乜嘢規律喺當中。例：用機械學習演算法做返咗個聚類分析，結果係消費者可以按照佢哋對件產品各方面嘅評價分做三大類。
匯報個結果（尤其係步驟 3 嘅結果）俾相關人員聽，將數據視覺化，等啲數據易睇。
講吓個結果有乜嘢影響－即係知道咗呢個結果，目前嘅做嘢方法要有乜嘢改變先可以更有效噉達到目的。例：知道咗消費者可以分做三類，產品設計要點改變先可以令消費者更加滿意？

睇埋

參考

Understanding The 4 V's Of Big Data, Forbes，呢篇文講到大數據有 4V，即係指數據量大（volume）、花款多（variety）、攞得快（velocity）、而且有時難核實（veracity）。
Cielen, D., Meysman, A., & Ali, M. (2016). Introducing data science: big data, machine learning, and more, using Python tools. Manning Publications Co.
Grus, J. (2019). Data science from scratch: first principles with python. O'Reilly Media.

引述

↑ Dhar, V. (2013). "Data science and prediction". Communications of the ACM. 56 (12): 64–73.
↑ Jeff Leek (12 December 2013). "The key word in "Data Science" is not Data, it is Science 互聯網檔案館嘅歸檔，歸檔日期2018年8月21號，.". Simply Statistics.
↑ Leskovec, Jure; Rajaraman, Anand; Ullman, Jeffrey David. Mining of Massive Datasets (PDF) (Preprint of 3rd ed.). Cambridge University Press. p. 1.
↑ Tansley, S., & Tolle, K. M. (2009). The fourth paradigm: data-intensive scientific discovery (Vol. 1). A. J. Hey (Ed.). Redmond, WA: Microsoft research.
↑ Hayashi, Chikio (1 January 1998). "What is Data Science? Fundamental Concepts and a Heuristic Example". In Hayashi, Chikio; Yajima, Keiji; Bock, Hans-Hermann; Ohsumi, Noboru; Tanaka, Yutaka; Baba, Yasumasa (eds.). Data Science, Classification, and Related Methods. Studies in Classification, Data Analysis, and Knowledge Organization. Springer Japan. pp. 40–51.
↑ Bell, G.; Hey, T.; Szalay, A. (2009). "COMPUTER SCIENCE: Beyond the Data Deluge". Science. 323 (5919): 1297–1298.
↑ "ASA Statement on the Role of Statistics in Data Science". AMSTATNEWS. American Statistical Association.
↑ A Design Thinking Mindset for Data Science. Towards Data Science

拎

（英文） Towards Data Science

[1] Dhar, V. (2013). "Data science and prediction". Communications of the ACM. 56 (12): 64–73.

[2] Jeff Leek (12 December 2013). "The key word in "Data Science" is not Data, it is Science 互聯網檔案館嘅歸檔，歸檔日期2018年8月21號，.". Simply Statistics.

[3] Leskovec, Jure; Rajaraman, Anand; Ullman, Jeffrey David. Mining of Massive Datasets (PDF) (Preprint of 3rd ed.). Cambridge University Press. p. 1.

[4] Tansley, S., & Tolle, K. M. (2009). The fourth paradigm: data-intensive scientific discovery (Vol. 1). A. J. Hey (Ed.). Redmond, WA: Microsoft research.

[5] Hayashi, Chikio (1 January 1998). "What is Data Science? Fundamental Concepts and a Heuristic Example". In Hayashi, Chikio; Yajima, Keiji; Bock, Hans-Hermann; Ohsumi, Noboru; Tanaka, Yutaka; Baba, Yasumasa (eds.). Data Science, Classification, and Related Methods. Studies in Classification, Data Analysis, and Knowledge Organization. Springer Japan. pp. 40–51.

[6] Bell, G.; Hey, T.; Szalay, A. (2009). "COMPUTER SCIENCE: Beyond the Data Deluge". Science. 323 (5919): 1297–1298.

[7] "ASA Statement on the Role of Statistics in Data Science". AMSTATNEWS. American Statistical Association.

[8] A Design Thinking Mindset for Data Science. Towards Data Science

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]