1. 1
臺北市公務人員訓練處-106年度數位學習人才培育班@2017/12/14。台北
T h i n k B i g f o r B i g D a t a - D a t a G o v
講 師 : 張 大 明 ric h a rd @a b c te c h . p ro
L i n e / W e c h a t I D : a s k 2 b i g r i c h
認識人工智慧與
發展應用
(含參訪課後感想)
—科技創新的價值與機會
The Economist
Google leads in the race to dominate artificial intelligence
-Tech giants are investing billions in a transformative technology
https://goo.gl/L1swyG
13. Top 10 Hot Artificial Intelligence (AI) Technologies
Natural Language
Generation
自然語言生成
Speech Recognition
語音識別
Virtual Agents
虛擬代理人
Machine Learning
Platforms
機器學習平台
AI-optimized
Hardware
AI優化的硬件
Decision
Management
決策管理
Deep Learning
Platforms
深度學習平台
Biometrics
生物識別技術
Robotic Process
Automation
機器人過程自動化
Text Analytics and
NLP
文本分析和NLP自然語
言處理
13
There is no defined business case 42%
Not clear what AI can be used for 39%
Don’t have the required skills 33%
Need first to invest in modernizing data mgt platform 29%
A Narrative Science survey found last year that 38% of enterprises are already using AI, growing to 62% by 2018. Forrester Research
predicted a greater than 300% increase in investment in artificial intelligence in 2017 compared with 2016. IDC estimated that the AI
market will grow from $8 billion in 2016 to more than $47 billion in 2020.
https://www.forbes.com/sites/gilpress/2017/01/23/top-10-hot-artificial-intelligence-ai-technologies/#709c25af1928
14. 14
The Economist
Google leads in the race to dominate artificial intelligence
-Tech giants are investing billions in a transformative technology
https://goo.gl/L1swyG
15. 15
The Economist
Google leads in the race to dominate artificial intelligence
-Tech giants are investing billions in a transformative technology
https://goo.gl/L1swyG
20. 20T h i n k B i g f o r B i g D a t a - D a t a G o v
World of Watson, these are IBM’s 4 pillars and Big Blue is gearing up for an arms race and a big war - the war for mind
share and for money. According to analyst firm IDC, big data and analytics spending will hit $203 billion by 2020.
https://www.linkedin.com/pulse/war-4-pillars-big-data-cognitive-cloud-iot-bill-hayduk/
74. • A Culture of Big Data: Case Studies from Google
• Data Science at Riot Games
• 善用資料改善線上教育 : 一個十五人團隊,如何從無到有,融入資訊做決策
• 用數據解決都會城市的停車問題
• 有志一同:社群網路分析、電商、零售及網路行銷
• Data as a Service: 數據的新經濟
• 隨機對照實驗在公共領域的應用
• 數據分析在運輸管理之應用
• 全球網路攝影機帶來的機會與挑戰
• 大數據下的情緒分析
• 一個賭徒的告白 2:交易策略建構與分析,為何你該賭小一點
• 銀行資料這樣玩
• 從薪酬制度讀 CEO 的行為心理學
• 數位廣告大數據
• How to Approach Data Science Problems from Start to End: Two Case Studies
• 自己的空氣品質自己量 : 談參與式環境感測的機會與挑戰
• 如何備料:資料的抓取、清理以及串接
• 心理與行為資料中的因與果
• 資料科學與防疫應用的結合 : 以登革熱防治為例
• 資料分析前的奏曲 : 談資料收集的挑戰
74
http://datasci.tw/agenda.php#!tab4
第一天 7/16
重要主題Big Data案例介紹
75. • 未來城市的任意門 (Mobility on Demand for Future Cities)
• 從電腦視覺看人工智慧 : 下一件大事
• 人類行為訊號處理 : 跨學科 (醫療、教育、心理) 應用實例分享、心得、展望
• 有志一同:資料視覺化、資訊安全、教育大數據、健康醫療、財務金融、人工智慧/機
器學習/深度學習、開放資料及個資保護
• 公私協力的公共服務 - 以資料面詮釋
• The Inception of Learning from Big Video Data
• 從手機解讀行為與心理
• Crowd Computing for Big and Deep AI
• 智慧型工程管考系統 : 資料分析經驗談
• 音樂資料檢索
• 當語音處理遇上深度學習
• 貓都學得會的手機維修資料分析
• The Growth of a Data Scientist
• Practical Issues in Machine Learning
• 沒有大數據怎麼辦 ? 會計師事務所的小數據科學
• Feature Engineering in Machine Learning
• 未來城市的交通運輸
• 漫談 R 的學習挑戰與 R 語言翻轉教室
• 娛樂產業中的資料科學家 : 談資料科學於線上遊戲與職業運動之應用
• 空間數據分析推動精準新農業
• Open Innovation as Strategic Plan
• 世代之爭爭什麼 ? 談談如何從調查資料挖掘出豐厚的意義
• 資料視覺化的 20 個小訣竅
75
http://datasci.tw/agenda.php#!tab5
第二天 7/17
重要主題Big Data案例介紹
104. The role of ICTs in the e-commerce
transaction/EDI value chain
104Source: OECD(2013/10)
貿易商 & 代理商
vs. 原廠
電子商務、跨境電商
、行動商務、第三方支付
電商平台 & 垂直電商
vs. 官網 (品牌)
117. 機器學習 Machine Learning
一個電腦程序從經驗E至學會,
針對某類任務T和性能指標P,
如果在T中的任務性能,
按P測量,提高了有經驗的E。
A computer program is said to learn from experience E with respect to
some class of tasks T and performance measure P if its performance at
tasks in T, as measured by P, improves with experience E.
117
Tom M. Mitchell, WIKI
另一個定義:機器學習目標是要電腦編碼運用舉例的資料,或過去的經驗來
解決已知問題。 麻省理工學院
https://en.wikipedia.org/wiki/Machine_learning
137. Impact
• TEDxBroadway演講,Ben Wellington 提供五項原則,
透過故事讓數據更有意義(Making data mean more
through storytelling)
– 1.連結人們的經驗 Connection with People's Experiences
– 2.專注一個創意 Focus One Idea
– 3.讓它簡單化 Keep it Simple
– 4.探索你最熟悉的 Explore the Things You Know Best
– 5.發生改變 Make an Impact
137
https://youtu.be/6xsvGYIxJok
154. 154
What
happened?
• 描述性分析
Descriptive
Analytics
Why did it
happen?
• 診斷式分析
Diagnostic
Analytics
What will
happen?
• 預測式分析
Predictive
Analytics
How can we
make it
happen?
• 指導性分析
Prescriptive
Analytics
價值
困難度 & 複雜度
傳統BI 進階分析-大數據
落後指標
洞見與覺察
領先指標
撿
找
人
算
神
算
Edited by Ta-Ming Chang , 2016/8/2
Big Data應用技術演變,
人 & 技術,大數據專案
本身
157. AlbiTer Mind Map
P1 設定目標
P2 擬定計畫
P3 闖關式教學設計
P4 實施闖關式教學
P5 產生倍力大數據
P6 持續螺旋式學習
Designed by Ta-Ming Chang & Roger Lo , 2015/9ABC.LSTP redoubles Big Data T+D Service Guide for Instructor
A l b i T e r ─ 推 動 學 教 移 轉 中 心 的 六 項 動 力
ABC.LSTP redoubles Big Data T+D Service Guide for Instructor
1
2
3
4
5
6
你感受過這樣的動力感覺嗎?
157
206. Open Data Top 20
1. Data.gov.uk the UK government’s open data portal including the British National Bibliography – metadata on
all UK books and publications since 1950.
2. Data.gov Search through 194,832 USA data sets about topics ranging from education to Agriculture.
3. US Census Bureau latest population, behaviour and economic data in the USA.
4. Socrata – software provider that works with governments to provide open data to the public, it also has its
own open data network to explore.
5. European Union Open Data Portal thousands of datasets about a broad range of topics in the European Union.
6. European Data Portal is a European portal that harvests metadata from public sector portals throughout
Europe. EDP therefore focuses on data made available by European countries. In addition, EDP also harvests
metadata from ODP.
7.DBpedia crowd sourced community trying to create a public database of all Wikipedia entries.
8. The New York Times a searchable archive of all New York Times articles from 1851 to today.
9. Dataportals.org datasets from all around the world collected in one place.
10. The World Factbook information prepared by the CIA about, what seems like, all of the countries of the world.
206
http://datatovalue.co.uk/top-20-open-data-sources/
207. Open Data Top 20
11. NHS Health and Social Care Information Centre data sets from the UK National Health Service.
12. Healthdata.gov detailed USA healthcare data covering loads of health related topics.
13. UNICEF statistics about the situation of children and women around the world.
14. World Health organisation statistics concerning nutrition, disease and health.
15. Amazon web services large repository of interesting data sets including the human genome project, NASA’s
database and an index of 5 billion web pages.
16. Google Public data explorer search through already mentioned and lesser known open data repositories.
17. Gapminder a collection of datasets from the World Health Organisation and World Bank covering economic,
medical and social statistics.
18.Google Trends analyse the shift of searches throughout the years.
19. Google Finance real-time finance data that goes back as far as 40 years.
20. UCI Machine Learning Repository a collection of databases for the machine learning community.
21.National Climatic Data Center world largest archive of climate data.
207
continue
236. TB級資料應用與分析才
能稱為大數據分析嗎?
Type A. Big data 問題跟 Small data 是一樣的
Type B. Big data 問題等同一大群 Small data 的問題
Type C. Big data 問題需要靠特製系統解決
Q9
http://readata.org/three-type-problem-for-big-data-analytics/