A presentation at Lecture by Shane Hudson
An Introduction To Artificial Intelligence It’s not all just robots! Discussion, what is AI? Identify AI challenges and non. Misconceptions.
A Few Questions • What do you think Artificial Intelligence is? • What is it used for? • Which AI systems have you used recently?
Alan Turing (23 June 1912 – 7 June 1954) Ask who recognises the photo.
Ask about things he is known for:
Enigma
Church Turing Thesis
Turing Machines
Turing Test
Alonzo Church (June 14, 1903 – August 11, 1995) Ask who this is, probably harder than previous.
Ask about things he is known for:
Church-Turing Thesis
Lambda Calculus
Undecidability of the Entscheidungsproblem
Turing’s doctoral advisor
Church Turing Thesis • In 1931, Kurt Gödel’s first incompleteness theorem proved that some true functions on integers cannot be computed. • The Church-Turing Thesis characterises exactly which functions can be computed. • It also states that a Turing machine is capable of computing any computable function.
Which Functions Are Not Computable? • Halting Problem
Proved by Alan Turing. • Decision Problem - “Is this statement provable, given true premises (axioms), using rules of logic?” Proved by Alonzo Church.
Turing Machines • An abstract idea, not a real physical machine. • Works with infinite storage, such as a infinite tape of binary digits. • A ‘head’ looks at a cell on the tape, it can both read and write, and move left or right. • The turing machine has defined states, which tell the head what to do with the cell it is looking at.
Turing Machine Diagram https://www.youtube.com/watch?v=gJQTFhkhwPA q 1 q 2 q 0 q H 1/0,L 0/1,· 0/0,L 1/1,R 0/0,· 1/0,L Initial Something that they can work on practically.
How It Works • 1/0,L - If 1, change to 0 and move left • 0/1,R - If 0, change to 1 and move right • 0/0,· - If 0, keep as 0 and don’t move
q 1 q 2 q 0 q H 1/0,L 0/1,· 0/0,L 1/1,R 0/0,· 1/0,L Initial 0 1 0 0 1 1 1 0 1 Halt Do this in groups on paper.
Turing Test
•
In his 1950, Alan Turing published a paper to Mind
journal titled ‘Computing Machinery and Intelligence’.
•
He discussed the idea of computers being able to think.
•
Most importantly, he proposed a test in which
somebody (or multiple people) have a conversation with
either a person or a machine without knowing which.
•
If there is a high success rate of thinking the machine is
a person, then the test is passed.
Oxford Dictionary “The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.” So, What Is Artificial Intelligence?
Some Categories of AI • Natural Computation • Computer Vision • Natural Language • Machine Learning • Simulation • Data Mining These are not official categories, because AI is such a loose term.
Neural Networks Talk about how each node is useless on its own. Discuss weighting etc.
Chess • It is unlikely that Chess will ever be solved, so you cannot know the correct move to make. Checkers was solved in 2007 and has far less moves. • On 11/5/1997, Deep Blue (invented by IBM) beat the then World Champion Garry Kasparov. • Deep Blue was able to compute 200 million positions per second.
Heuristics “Proceeding to a solution by trial and error or by rules that are only loosely defined” Oxford Dictionary
• Heuristics are at the heart of Artificial Intelligence. • An algorithm follows a set of rules. • A heuristic can be used to extend an algorithm by making an educated guess instead of following predefined rules.
University Essays “An essay on the biomedical applications of evolutionary algorithms for image segmentation on MRI scans of the brain” Explain the basics of Evolutionary Algorithms and the type of essays.
University Projects Explain that we can choose our own projects.
Have to persuade supervisors.
Okay, There Are Some Robots
• http://en.wikipedia.org/wiki/File:Alan_Turing_photo.jpg
• http://en.wikipedia.org/wiki/File:Alonzo_Church.jpg
• https://www.youtube.com/watch?v=gJQTFhkhwPA
• www.youtube.com/watch?v=dNRDvLACg5Q
• https://www.youtube.com/watch?v=I9hBeJQUFYg
• https://www.youtube.com/watch?v=aLpaLu2ZVOo
• http://tex.stackexchange.com/questions/132444/diagram-of-an-artificial-neural-network
• http://en.wikipedia.org/wiki/File:Falxmeningeom_MRT_T1_mit_Kontrastmittel.jpg
• http://www.telegraph.co.uk/motoring/news/8876686/Honda-Asimo-robot-advances-in-leaps-and- bounds.html
• https://www.plymouth.ac.uk/schools/school-of-computing-and-mathematics/robot-football
• http://i.dailymail.co.uk/i/pix/2015/02/09/2585F3CF00000578-0-image-a-2_1423524380512.jpg
References
This was a short lecture given as an introduction to AI and the kind of projects to expect at university.
The following resources were mentioned during the presentation or are useful additional information.
for free. You
can too.