automation, intelligence and knowledge modelling

40
Automation, intelligence and knowledge modelling Veselin Pizurica web11.org September 30th, 2016

Upload: veselin-pizurica

Post on 16-Apr-2017

293 views

Category:

Science


0 download

TRANSCRIPT

Page 1: Automation, intelligence and knowledge modelling

Automation, intelligence and knowledge modelling

Veselin Pizurica

web11.org September 30th, 2016

Page 2: Automation, intelligence and knowledge modelling

Veselin Pizurica CTO, [email protected]

Page 3: Automation, intelligence and knowledge modelling
Page 4: Automation, intelligence and knowledge modelling
Page 6: Automation, intelligence and knowledge modelling

So you graduated! How will you spend your time?

Page 7: Automation, intelligence and knowledge modelling

http://ml-games.tomasz-rewak.com/

Reinforced Learning

Page 8: Automation, intelligence and knowledge modelling

People are looking into Biology for inspiration, but things are sometimes a little bit other way around...

Page 10: Automation, intelligence and knowledge modelling

Machine vision, NLP, ...

Page 11: Automation, intelligence and knowledge modelling

Google: DeepMind to cut data center energy bills

http://www.theverge.com/2016/7/21/12246258/google-deepmind-ai-data-center-cooling

Page 12: Automation, intelligence and knowledge modelling

Deep Learning● Choosing the correct feature representation of input data, is a way that people can

bring prior knowledge of a domain to increase an algorithm's computational performance and accuracy.

● To move towards general artificial intelligence, algorithms need to be less dependent on this feature engineering and better learn to identify the explanatory factors of input data on their own.

● Deep learning tries to move in this direction by capturing a 'good' representation of input data by using compositions of non-linear transformations.

Page 13: Automation, intelligence and knowledge modelling

But there is a problem...

Page 14: Automation, intelligence and knowledge modelling

http://nautil.us/issue/40/learning/is-artificial-intelligence-permanently-inscrutable#disqus_thread

Pneumonia/Asthma story

Page 15: Automation, intelligence and knowledge modelling

DARPA: Modern learning algorithms show a tradeoff between human interpretability, or explainability, and their accuracy. Deep learning is both the most accurate and the least interpretable.

What vs. Why?

Page 16: Automation, intelligence and knowledge modelling

WHAT IS INTELLIGENCE?

Page 17: Automation, intelligence and knowledge modelling

Is he the Smartest Man in the World?

Page 18: Automation, intelligence and knowledge modelling

WHAT IS INTELLIGENCE?

Page 19: Automation, intelligence and knowledge modelling

WHAT IS INTELLIGENCE?

Page 20: Automation, intelligence and knowledge modelling

WHAT IS INTELLIGENCE?

Page 21: Automation, intelligence and knowledge modelling

Pick your book

Page 22: Automation, intelligence and knowledge modelling

Swarm Intelligence

Page 23: Automation, intelligence and knowledge modelling
Page 24: Automation, intelligence and knowledge modelling

Why are kids so annoying?

Page 25: Automation, intelligence and knowledge modelling

Why is ice slippery?

Page 26: Automation, intelligence and knowledge modelling

Why is ice slippery?

Page 27: Automation, intelligence and knowledge modelling

Y = f (X) Y = f (X)

Page 28: Automation, intelligence and knowledge modelling

Rule engine is a

knowledge modeling problem

Y = f (X)

Page 29: Automation, intelligence and knowledge modelling

Modelling problem

http://www.theatlantic.com/science/archive/2016/06/how-consciousness-evolved/485558/Attention Schema Theory (AST)

Page 30: Automation, intelligence and knowledge modelling

Decision Tree

Learn rules from dataApply each rule at each nodeClassification is at the leafs of the tree

Page 31: Automation, intelligence and knowledge modelling

Let’s do it!

How many distinct decision trees we have with n Boolean attributes?= number of distinct truth tables with 2^n rows = 2^n^nWith 6 Boolean attributes 18,446,744,073,709,551,616

Page 32: Automation, intelligence and knowledge modelling

Let’s talk about pancakes!

Page 33: Automation, intelligence and knowledge modelling

Bayes Nets

Page 34: Automation, intelligence and knowledge modelling

Let’s be Engineer!

Page 35: Automation, intelligence and knowledge modelling

How do you express that car needs both battery and fuel to function? Easy.

How do you say that if your lights are not working, most likely it is a battery fault, but it could be as well that just lights are broken? Still the fact that lights are not working point to most likely cause of the battery fault.

If you only model via composition and add behavior separately – what most of the tools do these days – you are heading for complexity!

Page 36: Automation, intelligence and knowledge modelling
Page 37: Automation, intelligence and knowledge modelling

But there is a problem...

Page 38: Automation, intelligence and knowledge modelling
Page 39: Automation, intelligence and knowledge modelling

http://web.mit.edu/cocosci/Papers/Science-2015-Lake-1332-8.pdf

Glenn Roberts and Motorcycle Mojo Magazine

Page 40: Automation, intelligence and knowledge modelling

Pick them all!