![Q&A How are you going to implement the projects? We are looking for partners who see great business opportunities in A.I. and Machine Learning spaces What kind of partners are you looking for? Strategic partners: corporations or investors from the U.S., European Union and Asia Contact us by email: [email_address] Valery Tsourikov 2011 (c). All rights reserved](https://image.slidesharecdn.com/aicreatesfinalupdate-110412055943-phpapp02/85/Ai-creates-final-update-27-320.jpg)
Ai creates final update
- 1. A.I.Creates Intellectual Capital Club Creating intellectual capital for partnerships in predictive analytics and artificial intelligence Dr. Valery Tsourikov Minsk 2011 tm Valery Tsourikov 2011 (c). All rights reserved
- 2. Part 1. Business Model: Slides 3-10 Part 2. Predictive Analytics, A.I. for content generation: Slides 11-26 Valery Tsourikov 2011 (c). All rights reserved
- 3. Modern Artificial Intelligence Means Big Business Machine Learning, Combinatorial Intelligence, Semantics, Predictive Analytics can generate : Better content Faster & Cheaper , than human team alone Business opportunity is really big. For example, over the past five years IBM spent $ 14B for 24 analytics acquisitions. How to exploit this mega-trend? Valery Tsourikov 2011 (c). All rights reserved
- 4. Gartner Group. Hot Technologies for 2011: N ext-Generation Analytics and Social Analytics Next-Generation Analytics : It is becoming possible to run real-time simulations of each individual business action to predict their future outcome rather than simply providing data about past interactions. Valery Tsourikov 2011 (c). All rights reserved
- 5. Idea of the Intellectual Capital Club Our Ph.D. teams develop A.I. software Joint venture is formed with strategic partner or investor JV exploits hot market opportunity Projects Robust trading systems Global markets prediction Credit scoring HFT platforms Drug discovery Business analytics Marketing trends Image and video mining Price prediction: insurance Valery Tsourikov 2011 (c). All rights reserved World Markets Joint Venture Strategic Partner/Investor A.I. Software Team
- 6. The Idea: To Create Critical Mass of Human Brains Who Will Code A.I. Systems In September 2010 we started special course on advanced analytics with a clear business objective: To teach best talents to develop A.I. systems for predictive analytics and automatic content generation So far, 65 mathematicians and programmers completed the course; many more new researchers will join the Club in 2011 Several new business projects in predictive analytics have already been initiated by course graduates Valery Tsourikov 2011 (c). All rights reserved
- 7. How the Club Works We train Ph.D. students in Predictive Analytics and A.I. and invite best graduates to start real software projects Then we show prototypes of intelligent systems to potential strategic partners/investors Together with partners we exploit hot business opportunities on world markets Valery Tsourikov 2011 (c). All rights reserved
- 8. Course Attendees Master and Ph.D. students majoring in: - mathematics artificial intelligence computer science econometrics computer linguistics financial engineering We invite only best Ph.D. students from five major universities in Minsk Valery Tsourikov 2011 (c). All rights reserved
- 9. Course Content A.I. and advanced analytics Mega-trend: content by A.I. Machine Learning - Trees, Kernels - Feature selection - Support Vector Machines - ML application Predictive Analytics A.I. and novelty creation - Building Blocks - Novelty Propagation - Semantics - Genetic Programming Drugs discovery with ML Pattern recognition Data, text, image mining Ensembles, boosting Causal time series Granger causality Extended Kalman filter Sequential Monte Carlo Automatic creation of mathematical models Trading strategies Credit scoring Price prediction High frequency trading Valery Tsourikov 2011 (c). All rights reserved
- 10. Ongoing Projects DISFA: automated creation of trading systems for ETFs and U.S. stocks. Shipping to clients since Q4’2010 Credit scoring: fast SVM creates credit risk models Started in Q1’2011 . Partner: AAI, U.S., markets - U.S. , China A.I. Platform for high frequency trading with Predictive trading strategies. Started in Q1’2011. Partner: DISFA Global Very High Probability Patterns: creation of new generation of trading systems and optimal portfolio models, which have very good steady profit and low draw down . Started in Q1’2011. Partner: Advanced Analytics International, U.S.A. Valery Tsourikov 2011 (c). All rights reserved
- 11. New Projects: Partners are Welcome Predictive engines to find causal relations between financial time series for global markets. Launch date March 31’2011 Business model : sell subscription to investment funds Technology : Sequential Monte-Carlo, Causal Reasoning Universal Analytics System with fast Support Vector Machine for very large data sets. Very broad market Valery Tsourikov 2011 (c). All rights reserved
- 12. Valery Tsourikov 2011 (c). All rights reserved A.I.Creates News Nov 23, 2010 New group is formed – to develop A.I. software for automatic creation of Predictive StatArb Trading Systems Team: Sergey - Stepan - Valery - Artyom Code name AWALI – Synthesis. Partner: DISFA Global
- 13. Valery Tsourikov 2011 (c). All rights reserved A.I.Creates News December 16, 2010 New group is formed – to develop A.I. software for finding causal relations between economic and financial time series Team: Igor-Alexey-Anatoly-Christina
- 14. Valery Tsourikov 2011 (c). All rights reserved A.I.Creates News March 31, 2011 New group is formed – Very High Probability Patterns, to reach very high probability of winning trades and very low draw down Team: Sergey – Margarita - Maxim - Catherine
- 15. Rapid R&D Methodology To minimize new products launch cycle we use a novel method called RRD – Rapid Research and Development – check presentation on the website Valery Tsourikov 2011 (c). All rights reserved
- 16. Part 2.Content Revolution Predictive Analytics and A.I. for content generation Valery Tsourikov 2011 (c). All rights reserved
- 17. Internet and Intranets 2011+: Major Shift to Content Internet evolution Phase 1: infrastructure (Sun Microsystems, Cisco), search engines (Yahoo, Google), e-shops (Amazon, EBay) Phase 2: user-generated simple content (Flickr, blogs, social media – Facebook, LinkedIn) Phase 3: - A.I. generates content New MEGA TREND Valery Tsourikov 2011 (c). All rights reserved
- 18. A . I . Creates Content as Mega-trend Addresses basic needs: content creation is fundamental feature of human civilization Internet is ready: everywhere, high bandwidth, cheap A.I. is ready: machine learning, pattern recognition, advanced analytics, semantics Huge business opportunities : decisions support; business analytics, robotic trading; data, images and video mining Valery Tsourikov 2011 (c). All rights reserved
- 19. Ready to explore new opportunities? Internet 2011+: Content Revolution Many groups will try to ride mega-trend Valery Tsourikov 2011 (c). All rights reserved
- 20. Why A.I. for content creation? A.I. silicon brain Very fast evolution Works 24x7x365 Silicon brain capacity is practically unlimited No salary increase, at least for now Human brain No change for 20,000 years Lazy, sleeps, drinks etc. Brain capacity is limited Demands salary increase, bonuses, vacations and respect no matter what A.I. generates content - Better - Faster - Cheaper Valery Tsourikov 2011 (c). All rights reserved
- 21. Artificial Intelligence vs. Human Brain Human brain stopped evolving 20 000 years ago Artificial Intelligence is entering exponential phase of evolution A.I. humans This is fundamental fact: A.I. evolves exponentially but human brain remains unchanged since Stone Age Valery Tsourikov 2011 (c). All rights reserved
- 22. A.I. Beats Human Garry Kasparov, the greatest chessplayer of all time , was defeated by Deep Blue computer in 1997 Same year, Invention Machine Corp. announced A.I. software that create d new patentable ideas by linking thousands of physical effects in a second No human can generate ideas with such speed Valery Tsourikov 2011 (c). All rights reserved
- 23. Why is power of A.I. rapidly increasing ? First, because microprocessors double their performance every 18 month (Moore’s law) Second, modern A.I. understands the meaning of words and can learn by reading different information sources A.I. is able to run Darwinian-type concept creation and selection with Monte-Carlo method Valery Tsourikov 2011 (c). All rights reserved
- 24. users Our Niche: Advanced A.I. for Content Generation A.I.+ Human Creativity! Valery Tsourikov 2011 (c). All rights reserved
- 25. Content Creation by A.I. Creative Artificial Intelligence Facts Trends Models Theories Needs Common sense Concepts News New content : Risk models Predictions Images Videos Discoveries Trading systems Optimal Portfolios Valery Tsourikov 2011 (c). All rights reserved
- 26. Machine Learning : Great Business Opportunities on Modern Internet Internet will be filled with Machine Learning algorithms Valery Tsourikov 2011 (c). All rights reserved Internet : texts images data video Machine Learning: Text Mining Image Recognition Predictive Analytics Video Search
- 27. Q&A How are you going to implement the projects? We are looking for partners who see great business opportunities in A.I. and Machine Learning spaces What kind of partners are you looking for? Strategic partners: corporations or investors from the U.S., European Union and Asia Contact us by email: [email_address] Valery Tsourikov 2011 (c). All rights reserved