Lecture Applied data science: Application

Lecture "Applied data science: Application" includes content: digital transformation, data science lifecycle, data science application,... We invite you to consult!
Nội dung trích xuất từ tài liệu:
Lecture Applied data science: Application Applied Data Science Sonpvh, 2022 1. Introduction 8. Validation 2. Application 9. Regularization 3. EDA 10. Clustering 4. Learning Process 11. Evaluation 5. Bias – Variance TradeOff 12. Deployment 6. Regression 13. Ethics 7. Classification 1 DIGITAL TRANSFORMATION PROCESS CURRENT INDUSTRY 4.0 BUSINESS 2 “87% of data science projects never make it into production”VentureBeat AI, 2019 PEOPLE THINK AI WILL [4] DESTROY THE WORLD, “85% of big data projects fail” Gartner, 2017 [5] AND ACTUAL IS … “Through 2022, only 20% of analytic insights will deliver business Gartner, 2019 [6] outcomes” 3 ADAPTAPILITY LACK OF TALENTs WRONG PROCESS 4 • Can you help me predict user demand? • I have this data, can you predict that label? • I want to sell milk for pregnant on ZingNews, can you predict pregnant one? • 4.0 Real Estate, 4.0 Bank, 4.0 research, 4.0 …. • I applied model on relevant product, why not work??? • Our AI is smarter than … 5 • We have computer vision, NLP, social analysis, AR, VR … on this Build their own platform 2025 2018 • ??? • Data & AI 2010 • Mobile - IoT Use the right platform 2000 • Internet - web Digital Transformation Digitalization Digitization 6 From digitization through digitalization, to digital transformation, Dobrica api-and-digitization-digitalization-and-digital-transformation , Amancio, 2019 [3] 7 Savic, 2019 [1] Target of digital transformati on 1. Data which 1. Smart 2. Privacy, 2.Profitable Regulatory could be utilized 3.Painpoint/ 3. Comercialize Added value Target of digital transformation Extra: Target of Data Business Roadmap to the digital transformation of business models Src: “Digital Transformation now! Guiding the Successful Digitalization of Your Business Model”, D. R. A. Schallmo, C. A. Williams, 2018 [8] 8 MARKET • Business Problems: Important • Data collection CUSTOMER • Label definition: Difficult COLLEAGU • Modeling • Label collection: Costly Es • Evaluation • Evaluation strategy: Need a lot of experiences • Deployment • Data collection & preparation: & Time $$$ • Modeling: 20% time – but everyone talk about this • Offline & Online Validation: Pray BOSS HOPE TEAM • Deployment: MEMBERS PROVIDER 9 Sale • Applied Data projects need a team!!! • You are changing the culture, not a solution Tech Product • It’s costly, and most time, it doesn’t work • But Fear Of Missing Out (FOMO) • The results are great, if success Marketi WELCOME TO APPLIED DATA SCIENCE Data • ... ng 10 1. From digitization through digitalization, to digital transformation, Dobrica Savic, 2019 2. D. R. A. Schallmo, C. A. Williams, Digital Transformation Now!, SpringerBriefs in Business, 2018 3. https://www.amanciobouza.com/post/api-and-digitization-digi ...