Teaching of Genetic Algorithm with Excel

In a previous post I discussed using Scratch and Excel to model neurones. This post looks at using Excel and six-sided dice as a way of developing insights into how  Genetic Algorithm work, before going on to program one. 

A very simplified version of Tournament Selection is used for the parent selection and the mutation works by rolling a die to get a number between 1-6.

The problem to be solved is to find the lowest values for x and y in the equation 
(x-6)*(x-6)+(y-1)*(y-1).

Routine

1. Using an Excel spreadsheet,  roll two dice six times. Fill in the first two columns with these numbers - these are X and Y values for each solution.
2. The fitness scores should be calculated based on the equation. Low values for this problem are best.
3. 1^st Parent: Roll two dice, if the numbers are same reroll one die to until the numbers are different. Use the two values to select the 1^st parent, the solution with the lowest fitness of the two. Take the X part of the selected parent and it forms the X part of the new child solutions.
4. 2^nd Parent: Roll two dice, if numbers are the same or appear in 1^st parent, reroll until you get two different numbers (including different to the 1^st parent). the solution with the lowest fitness of the two. Take the Y part of the selected parent and it forms the Y part of the new child solution.
5. Mutation: Roll a die for each part of the child solutions. If the roll is 1, roll another die and replace the appropriate element with the new number – even if the same as the previous value.
6. Copy the average  into the table and the lowest value 

7. Copy the child solutions after mutation (orange) into the yellow area and repeat steps 1-6 10 times 

All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon

BCS (Bedford) Artifical Intelligence: How did we get here and where are we going?

Taken from: http://beds.bcs.org/event.php?event=130

Artifical Intelligence: How did we get here and where are we going?

Date	Tuesday 21 Feb 2017
Time	18:30-20:00
Registration	Please register for this event
Location	Tavistock Suite, The Park Inn Hotel, 2 St Mary's Street, Bedford MK42 0AR
Speaker	Professor Max Braemer, University of Portsmouth.
Abstract	Professor Bramer joined the small Artificial Intelligence (AI) research community in Britain as a part-time PhD student in the early 1970s, just in time to see the field virtually wiped out in this country following a very hostile report to the main national funding agency. AI techniques have become near-ubiquitous in recent years, often under other names such as Recommendation Systems, Intelligent Robotics, Conversational Agents or Big Data. It is now possible to discuss the idea of a machine possibly becoming intelligent in respectable circles and AI has gone from being dismissed as of no value 40 years ago to being denounced today as a possible threat to the survival of our species. In this talk Professor Bramer will give a personal and highly selective view of the history of this important field and he will look at some of the ways in which AI is now in everyday use and some recent high-level applications, together with the issues of legal liability some of them raise. Finally he will consider some alternative possible AI futures and to what extent the warnings that are now being made about AI by some prominent people are justified. Max Bramer is Emeritus Professor of Information Technology at the University of Portsmouth. He is Chair of the British Computer Society Specialist Group on Artificial Intelligence (SGAI) and Vice-President of the International Federation for Information Processing (IFIP), a consortium of around 50 national and international IT societies, including the BCS and the ACM. He is also a Chartered Fellow of the BCS. Professor Bramer has been actively involved in the AI field since the early 1970s. He has led or supervised many AI projects since then, mainly in Knowledge Based Systems and the area that is now known as Data Mining or Big Data. He has written extensively on AI and other topics with over 200 publications, including many books and edited collections of papers. He has also presented well over 100 external lectures, seminars, etc. on a variety of topics, principally concerning Artificial Intelligence and the use of computers in education, including lectures in Denmark, Canada, Holland, Spain, Japan, the Republic of Ireland, the former Yugoslavia, France, Germany, Greece, Bulgaria, Poland, the United States, China and Australia. The third edition of his book 'Principles of Data Mining' is scheduled for publication in January 2017.
Agenda	18:30 Registration, refreshments and networking 19:00 Guest speaker - Professor Max Bramer 20:00 Opportunity to network and talk to the speaker
Downloads	Flyer

All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon

Environmental Award for Computing

Gary Hill and Scott Turner from the Computing subject team were, and still are part, of a team working on the Race to the Top Schools Coding Competition the other partners are Northamptonshire County Council, Worshipful Company of Information Technologists and Code Club. The project has been recognised by a Green Apple Award for Environmental Best Practice.

More details on the competition can be found at: https://computingnorthampton.blogspot.co.uk/2016/02/race-to-top-coding-competition.html

Other Partners
- Northampton County Council: 
http://www.northamptonshire.gov.uk/en/councilservices/Environ/carbon/Pages/default.aspx

- Worshipful Company of Information Technologists: 
https://www.wcit.org.uk/members/anon/new.html?destination=%2Findex.html

- Code Club: 
https://www.codeclub.org.uk/

All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon

A survey of low power techniques for efficient Network-on-Chip design

Two new papers

A survey of low power techniques for efficient Network-on-Chip design.
Opoku Agyeman, M. and Ofori-Attah, E. 
 In: 23rd IEEE Symposium on High Performance Computer Architecture (HPCA). USA: IEEE.

Abstract
Power consumption continues to be a challenge for designers as the complexity of NoC increases. The scaling down of technology towards the deep nanometer era will only cause an increase in the amount of power NoC components will consume. Therefore, low power design solution is one of the essential requirements of future NoC-based System-on-Chip (SoC) applications. Several techniques have been proposed over the years to improve the performance of the NoCs, trading-off power efficiency; particularly power hungry elements in NoC routers. Power dissipation can be reduced by optimizing the router elements, applying architecture saving techniques and communication links. This paper presents recent contributions and efficient saving techniques at the router, NoC architecture and Communication link level.

For more details: http://nectar.northampton.ac.uk/9086/




A survey of low power NoC design techniques. 
Opoku Agyeman, M. and Ofori-Attah, E.  
In: International Conference on High-Performance Embedded Architectures and Compilers (HiPEAC). Sweden: ACM. (Accepted)

Abstract
As we usher into the billion-transistor era, NoC which was once deemed as the solution is defecting due to high power consumption in its components. Several techniques have been proposed over the years to improve the performance of the NoCs, trading off power efficiency. However, low power design solution is one of the essential requirements of future NoC-based SoC applications. Power dissipation can be reduced by efficient routers, architecture saving techniques and communication links. This paper presents recent contributions and efficient saving techniques at the router, NoC architecture and Communication link level.

For more details A survey of low power NoC design techniques




All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon

BCS Northampton Event: Cyber Security Lecture

Cyber Security Lecture

Scott Bullock 

Cloud Trust Officer

Forcepoint 

Thursday 26th January 2017

7pm with 7:30 

Rm NW205  

University of Northampton

Global Cyber security leader Forcepoint released its 2017 Cyber security Predictions Report last month.

This year's report examines the increasing convergence of the technological and the physical worlds and the long term implications of this new digital ecosystem on organizations and institutions worldwide

Cyber security experts from Forcepoint and Raytheon collaborated to develop these predictions.

Scott Bullock Forcepoint's Cloud Trust Officer will present an overview of the predictions along with an analysis of how well Forcepoint performed last year. Scott is a member of Northampton BCS Group Committee.

Location

Rm 205

Newton Building

University of Northampton

NN2 6JB

All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon