●本課程是什麼？本課程的宗旨為培養企管系學生具備「行銷數據分析思維」的素養，目的是訓練學生成為「商務分析師」。本課程不會也不想訓練學生成為資科科學家或資料工程師！其實際目標是讓學生修畢一學期的課程之後，即有能力與資料科學家或資料工程師「對話」。課程會以提升學生對於程式邏輯的理解與跨領域溝通能力為主，期望學生將來能與程式設計師溝通、進而換位思考，以增加企業管理系學生畢業後的職場競爭力。
通常一位行銷人與一位資料科學家碰面，對話如下：
「我們最近有一筆資料，想要分析看看，是否能從當中找出一些能幫助行銷策略的訊息。嗯…分析技術最好夠尖端，夠前衛，能幫助我在客戶或長官面前的簡報。」
「你的意思是分析結果最好能展現消費者最新的動態，能依群體制訂個人化內容，並希望最終的圖表能看起來時髦又現代？」
「是的！沒錯！」
「好，沒問題！首先，你的資料是監督式還是非監督式？還是兩者均有可能？」
「……（行銷人一臉呆滯…）」
在選修過本課程之後，我們會希望學生將對話變成以下的方式（記得，經理人（企管系的訓練目的）的任務是「整合」，其中包含了「理解」對方的語言）。
「好，沒問題！首先，你的資料是監督式還是非監督式？還是兩者均有可能？」
「是監督式的資料，因為我們的需求非常清楚。」
「但是所有的資料都有標記嗎？如果不是的話，你們會提供嗎？」
「我們這邊會提供清楚的定義，但我們可以討論一下標記的方法？…」
但請切記的是：「能寫程式的人多不勝數。但只有少數人能做對心理學（與策略）的部分。而講到商業行為，正確的心理學（或管理學）使得一切都不一樣。」我們雖然會著墨於科技工具的使用，但永恆的重點仍然在「管理學中的科學原則」，一種可讓你/妳使用數十年的管理觀點。
●本課程不是什麼？（1）本課程並不提供那些你會在雜誌上看到的管理（普通）知識；（2）本課程不會教授的範圍：人工智慧相關領域與雲技術。
第一，這門課程所教授並不是你一般所想像的「管理」，而是「管理學」背後的「科學」，一種可以讓你使用長長久久的知識。第二，一般而言，目前可以辨視圖像或語音等資料的技術屬類神經網路與深度學習相關。本課程的重點放在數字型資料的分析，圖像、語音與文字型資料的分析將被放置在「人工智慧行銷學」課程中。雖然在業界應用中，絕大部分的大數據分析必須在「雲技術」上運行，但了解其背後的科學原理與其應用並不需要雲技術。為了降低同學們學習的門檻，本課程將不會談到任何雲技術。
●本課程的難易度設定：本課程授課教師的教育哲學觀為，一個系所所提供的必選修課程必須是「專業」課程，需要能訓練學生「思考」並同時提供予「專業知識」，行有餘力則再將上述過程賦予「意義」。換言之，對企管系（或管理學院）的學生來說，本課程具有一定的難度。如果以東海企管的必修課來比擬，大概介於「作業管理」與「管理科學」之間。雖然本課程是專門為了「企業管理學系」的同學而設計的，將會把所需要的「數學」與「程式」基礎降到最低，只為了教授資料科學的行銷應用。但就如那句老格言：「在岸上學不會游泳」。為了深刻地理解資料科學的原理，程式實務的訓練是重要的。
●本課程所需的Python程度。我們將使用Python做為分析與實作的工具。Python 是一種強大的直譯式高階程式語言，其設計哲學強調可讀、易用、簡潔。Python的應用非常廣泛，像是企業管理可能會涉及的範圍，Python都可做為應用工具。像是對企業的資料與文字處理與分析、對複雜數據進行資料探勘等，幾乎各方面皆可使用Python。而且Python簡單學習且容易使用，相關的學習資源非常豐富，非常適合非資訊相關科系的同學學習與使用。
本課程的內容著重在教導同學了解Python 3的基本邏輯，熟悉Python開發工具（Spyder）的操作，並可以撰寫簡單的Python程式，用以分析數位行銷相關資料（尤其是那些無法使用Microsoft Excel分析的數據）。但基於考量企業管理學系的同學們學習Python（或其他程式）的時間不長，及同學們的實際需要，本課程教授的方法優先注重同學們在學習Python時的應用（尤其在行銷上的應用）。基於上述考量，雖然Python當中有許多實用的技巧（例如物件），均極富教學價值，但都牽涉及許多的進階的程式寫作（例如物件導向觀念），學習門檻較高，故均忍痛捨棄。本課程一開始將複習Python 3的基本程式設計能力，並快速簡短地教授matplotlib等繪圖技巧、而在學期的1/3左右開始進行銷（商業）資料分析。●What is this course? The purpose of this course is to train students in the business management department to have the "market data analysis thinking" and the purpose is to train and generate as a "business analyst". This course will not and do not want to train as a data scientist or data engineer! In fact, the actual goal is to allow students to "talk" with data scientists or data engineers after a one-year course. The course will focus on improving students' understanding of programming logic and cross-domain communication skills. It is expected that students will be able to communicate with program designers and then think about their positions in order to increase their field competition after graduation from the Department of Enterprise Management.
Usually a marketing person meets a data scientist and the conversation is as follows:
"We recently had a piece of information that we wanted to analyze and see if we could find out some information that could help marketing strategies. Well...Analytical technology is best to be sophisticated and easy to get ahead of me, and it can help me report in front of customers or bosses."
"You mean the analysis results should best show the latest consumer movements, to order personalized content based on groups, and hope that the final chart will look fashionable and modern?"
"Yes! That's right!"
"Okay, no problem! First of all, is your information supervised or non-supervised? Or is it possible for both?"
"… (The marketing man is stunned...)"
After choosing this course, we will hope that students will change their conversations in the following ways (remember that the task of the manager (the training purpose of the Department of Business Management) is "integration" that includes "understanding" the other party's language).
"Okay, no problem! First of all, is your information supervised or non-supervised? Or is it possible for both?"
"It's supervisory information because our needs are very clear."
"But are all the data tagged? If not, will you provide it?"
"We will provide clear definitions here, but can we discuss how to mark it?"
But please remember: "Most people who can write programs are not compulsory. But only a few people can do the psychology (and strategy). When it comes to business behavior, correct psychology (or management) makes everything different." Although we are obsessed with the use of technological tools, the permanent focus is still on "scientific principles in management", a management view that you/you have used for decades.
●What is this course not? (1) This course does not provide management (ordinary) knowledge that you will see on magazines; (2) The scope of this course will not be taught: artificial intelligence-related fields and cloud technology.
First, the teaching of this course is not the "management" you generally imagine, but the "science" behind "management", a kind of knowledge that can allow you to use for a long time. Second, generally speaking, the technical neural network that can currently distinguish data such as images or voice is related to in-depth learning. The focus of this course is on the analysis of digital data, and the analysis of image, voice and text data will be placed in the "Artificial Intelligence Marketing" course. Although most large data analysis must be operated in "cloud technology" in industry applications, cloud technology is not required to understand the scientific principles behind it and its application. In order to reduce the learning time of students, this course will not talk about any cloud technology.
●The difficulty setting of this course: The educational philosophy of the course taught by the teacher is that the required course provided by a department must be a "professional" course, which requires that students be able to "think" and provide "professional knowledge" at the same time. If you have the effort, you will then give the above process "meaning". In other words, this course has certain difficulties for students in the Department of Business Management (or School of Management). If compared with the compulsory courses of Donghai Enterprise Management, it is probably between "operation management" and "management science". Although this course is designed specifically for students in the "Enterprise Management Department", it will minimize the required "mathematics" and "programmatics" basics, and only teach marketing applications of data science. But just like the old saying: "You can't swim when you learn on the shore." In order to deeply understand the principles of data science, training of program practice is important.
●The level of Python required for this course. We will use Python as a tool for analysis and implementation. Python is a powerful direct-based high-level programming language, with a design philosophy that emphasizes readability, ease of use and simplicity. Python applications are very widespread. For example, the scope of enterprise management may involve, Python can be used as an application tool. Python can be used in almost all aspects, such as data processing and analysis of enterprise data and text processing and analysis, data exploration of complex data. Moreover, Python is simple and easy to use, with rich related learning resources, which are very suitable for students in non-information-related subjects.
The content of this course focuses on teaching students to understand the basic logic of Python 3, be familiar with the operations of Python development tools (Spyder), and write simple Python programs to analyze digital marketing-related data (especially those data that cannot be analyzed using Microsoft Excel). However, considering that students in the Department of Enterprise Management have not been learning Python (or other programs) for a long time and their actual needs, the method taught in this course focuses on the application of students when learning Python (especially in marketing). Based on the above considerations, although Python has many practical techniques (such as objects) that are very valuable for teaching, they all involve many advanced program writing (such as object-oriented concepts), and their learning doors are high, so they all endure the pain and give up. At the beginning of this course, we will refine the basic program design capabilities of Python 3, and quickly and briefly teach matplotlib and other drawing skills, and start marketing (business) data analysis around 1/3 of the learning period.
本課程是專門為了「企業管理學系」的同學而設計的。雖然Python的高階編程十分精妙且複雜，更可以實作像是資料結構或演算法等偏資工的科目，但為符合本課程的定位，本課程一律只使用簡單易懂的中文書（版），不會涉及任何較偏資訊工程的內容。
目前市面上並沒有一本教科書可以涵蓋Python的所有理論、軟體實作，通常需要大量閱讀書藉，而取各家所長。關於各課本是否需要購買及其使用情況，教師會在課堂上講說。對同學來說，除了必備的書藉之外，大部分的書藉在需要時，均可在圖書館借到（或在網路上查到）。
必備:
Foster Provost & Tom Fawcett（2016）。《資料科學的商業運用》。碁峯
孫宏明（2017）。《輕鬆學Python 3》。碁峯
中久喜健司（2018）。《Python 資料運算與分析實戰》。旗標
Python推薦閱讀：
Naomi Ceder（2019）。《Python 技術者們：練功！老手帶路教你精通正宗 Python 程式》。旗標
Allen B. Downey（2016）。《Think Python：學習程式設計的思考概念（第二版）》。歐萊禮
《精通 Python：運用簡單的套件進行現代運算》。碁峯出版。2017
Joel Grus（2016）。《Data Science from Scratch 中文版：用Python學資料科學》。碁峯
唐亘（2018）。《還在漫無頭緒？一本書帶你走完Python深度學習》。佳魁數位 （註：數學多。）
有賀康顕等（2018）。《機器學習：工作現場的評估、導入與實作》。碁峰
管理類推薦閱讀：
麥爾荀伯格、庫基耶（2013）。《大數據》。天下文化
馬汀．林斯壯（2017）。《小數據獵人》。寶鼎
城田真琴（2013）。《Big Data大數據的獲利模式：圖解．案例．策略．實戰》
岡（山鳥） 裕史（2010）。《從資料中挖金礦：找到你的獲利處方籤》。經濟新潮社
佩德羅．多明戈斯（2016）。《大演算》。三采
奈特．席佛（2013）。《精準預測》。三采
This course is designed specifically for students in the "College Management Department". Although Python's high-level programming is very sophisticated and complex, it can also be used as subjects that are biased in labor such as data structures or algorithms, in order to meet the positioning of this course, this course only uses simple and easy-to-understand Chinese books (versions) and will not involve any content of biased in information projects.
There is currently no textbook on the market that can cover all the theories and software implementations of Python. It usually requires a lot of reading and the seniority of each school. The teacher will talk about whether each course book needs to be purchased and used. For students, in addition to the necessary books, most books can be borrowed in the library (or found online) when needed.
Must be:
Foster Provost & Tom Fawcett (2016). "Business Usage of Data Science". Osho
Sun Hongming (2017). "Lightly Learn Python 3". Osho
Kanji Nakakuhika (2018). "Python Data Computing and Analysis" Flag mark
Python Recommended Reading:
Naomi Ceder (2019). "Python Technicians: Practice! The old hand-telling guide teaches you how to master authentic Python programs. Flag mark
Allen B. Downey (2016). "Think Python: Learning the Concept of Thinking Programming Design (Second Edition). Ole Legion
"Proficient in Python: Use simple suites for modern computing." Published by Osho. 2017
Joel Grus (2016). "Data Science from Scratch Chinese version: Learning Data Science in Python". Osho
Tang Geng (2018). "Is it still in a hurry? A book takes you through Python in-depth learning. Jiakui Digital (Note: Many mathematics.)
There are Casang et al. (2018). "Machine Learning: Evaluation, Introduction and Implementation of Work Fields". Grey Feng
Management recommendation reading:
Melzunberg, Kuchier (2013). "Big Data". World Culture
Matin． Linssie (2017). "Female Knob". Baoding
Makoto Shirota (2013). "Big Data's profit model: diagram. Case. Strategy. Real war》
Oka (Yamabi) Yushi (2010). "Mining gold mines from data: Find your profit label". Economic Trend Society
Pedro. Domingos (2016). "Big Calculation". Three Cai
Knight. Shifo (2013). "Precise Prediction". Three Cai
| 評分項目 Grading Method | 配分比例 Grading percentage | 說明 Description |
|---|---|---|
| 個人作業個人作業 Personal action |
20 | |
| 個人Python 小考個人Python 小考 Personal Python exam |
20 | |
| 團體上台簡報團體上台簡報 Group's briefing on the stage |
10 | |
| 團體期末報告團體期末報告 Group final report |
20 |
