Jorge Silva, Sr. Research Statistician Developer, SAS at MLconf ATL - 9/18/15MLconf
Estimating the Number of Clusters in Big Data with the Aligned Box Criterion: Finding the number, k, of clusters in a dataset is a fundamental problem in unsupervised learning. It is also an important business problem, e.g. in market segmentation. Existing approaches include the silhouette measure, the gap statistic and Dirichlet process clustering. For thirty years SAS procedures have included the option of using the cubic clustering criterion (CCC) to estimate k. While CCC remains competitive, we propose a significant and original improvement, referred to herein as the aligned box criterion (ABC). Like CCC, ABC is based on a hypothesis-testing framework, but instead of a heuristic measure we use data-adaptive reference distributions to generate more realistic null hypotheses in a scalable and easily parallelizable manner. We have implemented ABC using SAS’ High Performance Analytics platform, and achieve state-of-the-art accuracy in the estimation of k.
Jorge Silva, Sr. Research Statistician Developer, SAS at MLconf ATL - 9/18/15MLconf
Estimating the Number of Clusters in Big Data with the Aligned Box Criterion: Finding the number, k, of clusters in a dataset is a fundamental problem in unsupervised learning. It is also an important business problem, e.g. in market segmentation. Existing approaches include the silhouette measure, the gap statistic and Dirichlet process clustering. For thirty years SAS procedures have included the option of using the cubic clustering criterion (CCC) to estimate k. While CCC remains competitive, we propose a significant and original improvement, referred to herein as the aligned box criterion (ABC). Like CCC, ABC is based on a hypothesis-testing framework, but instead of a heuristic measure we use data-adaptive reference distributions to generate more realistic null hypotheses in a scalable and easily parallelizable manner. We have implemented ABC using SAS’ High Performance Analytics platform, and achieve state-of-the-art accuracy in the estimation of k.
Business DNA Model, Balanced Scorecard, and Strategy Map: A Visual Mathematic...Rod King, Ph.D.
This presentation features a 1-Page Diagram of the Business DNA Model as a platform for visually documenting, organizing, managing, and evaluating ideas on business models. The Business DNA Model can also be used to more deeply understood tools of Performance Management such as the Balanced Scorecard and Strategy Map as well as business modeling tools such as the Business Model Yacht, Business Model Canvas, and Lean Canvas.
http://goo.gl/qRZhwV
12 Disruption Vulnerabilities of the Business Model Canvas: BUSINESS MODEL CA...Rod King, Ph.D.
This presentation presents 12 "Disruption Vulnerabilities" or Achilles's Heels of the Business Model Canvas. Although the Business Model Canvas serves as a good tool for visually documenting a business model, it is limited in many respects especially with documenting, analyzing, and designing two/multisided markets (platforms). The tool of the Business Model Strip is presented as an alternative that eliminates the Disruption Vulnerabilities of the Business Model Canvas.
The Business Model Strip is designed with a multilevel paradigm so that it can be presented at various levels and in different visual formats. This presentation features the Business Model Strip in "canvas" (tessellation) format with 5 blocks (meso-level) as well as 9/11 blocks (micro-level). Finally, a visual template and checklist for an Exponential Business Canvas are presented.
1. 精準式影音行銷:廣告、短片、微電影
Revenue will follow viwers
人潮在那,營收就在那
Video is major trend for Internet
consumptions. But:
• Too many choices 選擇太多
• Less time to watch 時間有限
• High expectations 期待很高
3. 趨勢
案例
研究
分析
關鍵
要素
Cisco: over 90% Internet consumptions
商業
轉換
管理
需求
come from videos by 2015
The number of people watching
video on the Internet is expected to
nearly double by 2015 to 1.5 billion
5. 趨勢
分析
線上影音行銷的關鍵要素
案例 關鍵
研究 要素
商業 管理
轉換 需求
2. 抓位使用者的移動性 3. 創造影音被發現的機會
• SEO?
• Youtube?
• Content Curation
Collecting contents
that have value and
benefits to your
customers.
What can do, can not do between:
• iOS
• Android
• PC, Tablet and Smart Phone (Apps)
• OTT > Smart TV
8. 案例
趨勢
分析
關鍵
Online Video Platform
研究 要素
A comprehensive, scalable and flexible
商業 管理
轉換 需求
solution for the enterprise
管理影內容 發佈影音內容 滿足使用者體驗 使用行為及收視調查 收視轉換營收
Managing Content Publishing Content Creating Compelling Measuring Audiences Monetizing Video
User Experiences
S imp lify live s tream S treamline Rap id ly emb ed analytics P owerful tools to track
acq uis ition authenticated d elivery Op en P layer mod ules in p layer what’working, in real time
s
Framework enab les
Integrate w/exis ting Univers al content rap id ap p lication Centralize vid eo To refine your b rand ed
content, vid eo CMS d is trib ution d evelop ment analytics into s ingle content p rod uction,
Cros s -d evice d elivery d as hb oard tracking ad s d is trib ution and
Automated content and content monetis ation s trategy
to p hones , tab lets , recommend ations
game cons oles and Integrate real-time us er Integrate your own ‘
in
connected TVs S eamles s s ocial d ata hous e’ 3rd p arty
or
s haring ad vertis ing and
S up erior high b it rate
s treaming s p ons ors hip
Drive increas ed s ales
14. 趨勢
案例
研究
分析
關鍵
要素
Live Show / Social Network
商業
轉換
管理
需求
Product Demos / Global Strategies
15. Concepts of Video + CRM
Concepts:
• Interactive videos 互動式影音
• Get customer data via video
watching behavior & interactive domain social sharing geography
survey
結合會員資料及收視行為分析
• Segmentations 分眾精準分析
• Gender with model or
feature preference unique visitors external publishing devices
• Age with model or feature
preference
• Which countries or cities
have higher viewing rate
• Increasing action turning rate
(conversation rate)
提高行動發生轉換率
• What kind of video has the
highest action turning rate
• Places and video effectiveness (A,
B testing)
影音式新規格需求調查分析 Age: 18-24
比較性分析 A – B testing
Age: 12-18