Why 'Microsoft Predicts' predicted the US election so wrong?
Why & What is Big data?
What's 'machine learning' and why 'deep learning' is better?
How could 'deep learning' do for us?
The Paradigm Shift of IoT?
分成幾個部分:
1. 大數據與小數據(以美國總統選舉為例)
2. 大數據、機器學習與深度學習的關係
3. 大數據與數位廣告
4. 深度學習的應用
5. 美中台的 AI 發展
6. 人工智能+物聯網與數位典範翻轉
1. AI and Its Revolution
Craig Chao
Big Data Consultant & Senior Data Scientist
chaocraig@gmail.com
Exploring the AI development in
USA/China/TW
2. Agenda
u Let’s Start from United States Presidential Election,
2016
u Big Data - 4V
u Introduction of AI
u The AI Development in USA, China and Taiwan
u From Things to AIoT -- Reverse business
13. The Revolution of Big Data
DATA
Hypotheses
Statistical Analysis
BIG DATA
Hypotheses
Machine Learning
Data Mining
Machine-generated
Sampling, Multi-variant… All, Hyper space, …
Volume, Velocity, Variety, Veracity
Human-explainable
14. Models ßà Cases
Russ Merz, An Integrated Model of Media Satisfaction and Engagement: Theory, Empirical
Assessment and Managerial Implications, Journal of Applied Marketing Theory, Nov 2011
BIG DATA
Hypotheses
Machine Learning
Data Mining
Machine-generated
All, Hyper space, …
Volume, Velocity, Variety, Veracity
deductive inductive
Cases
Models
Models
Cases
15. Models ßà Cases
Russ Merz, An Integrated Model of Media Satisfaction and Engagement: Theory, Empirical
Assessment and Managerial Implications, Journal of Applied Marketing Theory, Nov 2011
19. Why Big Data?
u Internet Data
u 4 V
u Analysis àInference àAction
u Data-driven with no people interception
u Building model programmatically
u Hyperparameters optimization
Automation? Personalization?
24. 4R: Reach, Richness, Representation, Range
Reach
Richness
High
High
Low
使用者接觸量(DAU)
資料豐富度
(Behavioral data)
Range
High
系統範圍
( Affiliate of
whole context)
Representation
呈現形式與內容
(Format & Content)
25. Data Economy
Traditional -> Internet Economy
HighREACH
RICHNESS
High
Low
Traditional
Economy
Internet Economy
(quality)
(quantity)
聚焦、精準
擴大受眾
26. Reach: The Value Funnel
CPM campaign:
Revenue = N/1000 ⋅CPM
CPC campaign:
Revenue = N ⋅ CTR ⋅ CPC
CPA campaign:
Revenue = N ⋅ CTR ⋅
CVR⋅ CPA
UU Reach (DAU)
ARPU = Life-time Value
29. Richness
u Data Quality Richness
u Profile vs. Transaction
u Data Utilization Richness
u Call taxi (short vs. long route)
u Download times vs. Activation days
u Data Model Richness
37. High
4R: Reach, Richness, Representation, Range
Reach
Richness
High
High
Low 使用者接觸量(DAU)
資料豐富度
(Behavioral data)
Range
系統範圍
( Affiliate of
whole context)
Representation
呈現形式與內容
(Format & Content)
49. Matrix = Associations
Rose Navy Olive
Alice 0 +4 0
Bob 0 0 +2
Carol -1 0 -2
Dave +3 0 0
u Things are associated
Like people to colors
u Associations have
strengths
Like preferences and
dislikes
u Can quantify
associations
Alice loves navy = +4,
Carol dislikes olive = -
2
u We don’t know all
associations
Many implicit zeroes
Source: Sean Owen(2012), Cloudera
50. In Terms of Few Features
u Can explain associations by appealing to underlying features in
common (e.g. “blue-ness”)
u Relatively few (one “blue-ness”, but many shades)
(Alice)
(Blue)
(Navy)
Source: Sean Owen(2012), Cloudera
51. Losing Information is Helpful
u When k (= features) is small, information is lost
u Factorization is approximate
(Alice appears to like blue-ish periwinkle too)
(Alice)
(Blue)
(Navy)
(Periwinkle)
Source: Sean Owen(2012), Cloudera
60. AI Timeline
Ada
(1842)
Alan
Turing
(1950)
The first
conference
on
AI by John
McCarthy,
Marvin
Minsky
(1956)
Demonstrated
by Newell
(1957)
Unimation
s working
on GE
(1961)
Joseph
Weizenba
um (1965),
E.
Geigenba
um (1965)
Chess-
playing
program
by
Greenblatt
at MIT
(1968)
Jack
Myers
Harry
Pople
(1979)
1980s Ian
Horswil
l
(1993)
TiVo
Suggestions
(2005)
Apple,
Google,
Micorsoft
(2011)
Machine
Learning,
Deep
Learning
(2013 ~)
96. "I think it's a young girl throwing an
orange Frisbee in the park," Microsoft's
AI will tell you.
"I think it's a man jumping through the air doing a
trick on a skateboard," Microsoft's AI says.
160. AI investments
u 今年Google執行長 Sundar Pichai 在財報會議上明確表示,
DeepMind團隊的人工智慧技術幫助Google數據中心節省了
40%的能源利用
u 九月出席台灣半導體產業協會(TSIA)的聯發科創辦人蔡
明介就秀出數據
u 美國在自動駕駛車投資40億美元,無人機投資46億美元
u 日本產業經濟省對自動駕駛、機器人及產業保安等預計投入
145億美元
u 韓國規畫於2016年至2020年投入30億美元發展,建立國家級
AI中心
u 歐盟在2013年就啟動10億歐元「人類腦計畫」,同年,美
國宣布啟動45億美元的「腦計畫」
Src: http://www.bnext.com.tw/article/41689/ai-asian-silicon-valley-startups?utm_source=pnn
161. AI in Taiwan
u 科技部指出,政府用於補助與人工智慧技術研發如
深度學習、人工智慧、機器學習、無人機、自動駕
駛、機器人等相關6種領域,共1,691件計畫,總經
費約近13.65億元。
u 民間
u Appier在種子輪、A輪、B輪募資總計已獲得3千萬美
元(約10億元台幣)
u 創意引晴(Viscovery)在十月則獲得1千萬美元融資,
其團隊花了兩年研發出影音大數據分析技術,運用
人臉、圖片/商標、文字等七種不同的辨識技術,透
過深度學習計算影片情境及廣告投放效益。
u 盾心科技(Umbo CV)與行品科技(Skywatch
Innovation) 等公司也持續在AI領域深耕茁壯
Src: http://www.bnext.com.tw/article/41689/ai-asian-silicon-valley-startups?utm_source=pnn
174. Paradigm Shift(Reverse)
u Move
u Data à program
u Value
u Things à Product Service à Personal Service
u Value/revenue shift
u What if phone price is near its cost or free?