New trends in ICT: AI, agents, genetics & ubiquitous computing

Trends and Future Dir.
in ICT
Agent-Based Systems
Genetic Algorithm
Ubiquitous Computing
New trends and future directions of ICT

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.
28 May 2014
New trends and future directions of ICT, Slide 1
Trends and Future Dir. in ICT
Agent-Based Systems
Genetic Algorithm
Session Overview
1

Intelligent and Emotional Computing
Artificial Intelligence
Kansei Systems
Example in Kansei Engineering
Man-Machine Coexistence
Agent-Based Systems
New Challenges for Computer Systems
Characteristics of Agents
Multi-Agent Systems
Applications of Agents
Genetic Algorithm
Biological Exploration
Algorithm and Examples in GA
Application and Technology
Agent-Based Systems
Genetic Algorithm

Emotional Computing
Emotional Computing, What and Why?

People talking back to a computer/smart-phone is common
enough usually in a moment of frustration.
Getting the computer to respond in kind is a far different task
The challenge is not of inventing new software or hardware,
but to have ethics involved.
If computers are to have emotional components, what role
would they play in everyday life?
Do human beings really want an emotional relationship with a
mechanical mind?
The field is called Affective Technology
Agent-Based Systems
Genetic Algorithm
Affective Technology
Face Recognition
Today, machine prototypes exist that measure human

emotional expression through physiological signals such as
facial expressions and voice changes and allow a human-like
response.
Face Recognition and Applications
Agent-Based Systems
Genetic Algorithm
Affective Technology
Voice Recognition
Today, machine prototypes exist that measure human

emotional expression through physiological signals such as
facial expressions and voice changes and allow a human-like
response.
Voice Recognition and Applications
Agent-Based Systems
Genetic Algorithm

Intelligent Computing and Artificial Intelligence
Intelligence Computing
What is intelligence? Hard to define unless you list some
important characteristics, such as Reasoning, Learning, and
Adaptivity
Machine intelligence is: computer which follows problem
solving processes something like that in humans
Intelligent systems display machine-level intelligence,
reasoning, often learning, and self-adapting
Artificial Intelligence (AI) is usually defined as the science of
making computers do things that require intelligence when
done by humans.
Some important AI Techniques are: Neural Network, Genetic
Algorithm, and Expert Systems
Agent-Based Systems
Genetic Algorithm
Neural Network
Definition (Neural Network)

Neural Network are computational models inspired by an animals
central nervous systems (in particular the brain) which is capable
of machine learning as well as pattern recognition.
Artificial neural networks are generally presented as systems of interconnected neurons which can
compute values from inputs.
Applications are in Finger Print Recognition, Face Recognition, Patten Identification,etc.
Agent-Based Systems
Genetic Algorithm
Genetic Algorithm
Definition (Genetic Algorithm)

Genetic Algorithms (GAs) are adaptive heuristic search algorithm
premised on the evolutionary ideas of natural selection and genetic.
The basic concept of GAs is designed to simulate processes in natural system necessary for evolution,
specifically those that follow the principles first laid down by Charles Darwin of survival of the fittest.
As such they represent an intelligent exploitation of a random search within a defined search space to solve
a problem.
Applications are circuit/car design (http://rednuht.org/genetic_cars_2/), traffic controlling, path
finding, etc.
Agent-Based Systems
Genetic Algorithm
Expert Systems
Definition (Expert Systems)

An expert system is computer software that attempts to act like a
human expert on a particular subject area.
An expert system is made up of three parts:
A user interface - This is the system that allows a non-expert
user to query (question) the expert system, and to receive
advice.
A knowledge base - This is a collection of facts and rules. The
knowledge base is created from information provided by human
experts
An inference engine - This acts rather like a search engine,
examining the knowledge base for information that matches
the users query
Applications are Medical diagnosis, financial advice, discover locations to drill for water, vacation
advisor(http://www.exsys.com/demomain.html) etc.
Agent-Based Systems
Genetic Algorithm
Artificial Intelligence Software vs. Conventional Software

Application of Expert System
Artificial Intelligence Software vs. Conventional Software

Conventional computer software follow a logical series of steps
to reach a conclusion
Computer programmers originally designed software that
accomplished tasks by completing algorithms
AI software uses the techniques of search and pattern matching
Programmers design AI software to give the computer only the
problem, not the steps necessary to solve it
Agent-Based Systems
Genetic Algorithm
Our Attempt To Build Models Of Ourselves
Artificial Intelligence (Pros)

Ability to simulate human behavior and cognitive processes
(Intuition, Common sense, Judgment, Creativity, Beliefs etc)
Capture and preserve human expertise
Fast Response.
The ability to comprehend large amounts of data quickly.
Artificial Intelligence (Cons)

No common sense
Cannot readily deal with mixed knowledge
May have high development costs
Raise legal and ethical concerns
Agent-Based Systems
Genetic Algorithm
Kansei Systems
Definition (Kansei Engineering)
An consumer oriented technology for new product development
based on consumer Kansei (psychological image and feeling)
The attempt is to improve the business by producing product
or delivering services that fit to consumer feelings, emotions,
culture, profession, etc.
Agent-Based Systems
Genetic Algorithm
Kansei Systems
Attractive things Work Better
What is attractive? It is three level of processing

Reflective: Cerebral(use brains instead of hearts), emotive
(expressing a persons feelings)
Behavioral: Common, and interact with the world
Visceral : relating to deep inward feelings rather than to the
intellect
Affective designs impact on usability?

Working with the three levels
Visceral design : Appearance
Behavioral : Pleasure and effectiveness of use
Reflective: Self esteem and status
Appearance(s)
!Symmetry = Beauty = Usability ???
Fallll 2
Fa
2013
3
PSYCH
H / CS 6750
Appearance(s)
!Asymmetry = usable ==> beautiful ???
Fallll 2
Fa
201
2013
013
01
3
PSYCH / CS 6750
More!
!Asymmetry => usability (but, sexy?)
Fallll 2
Fa
201
2013
013
01
3
PSYCH / CS 6750
Behavior ?
!Feel => Function ??
Fall 2013
PSYCH / CS 6750
Reflective Design
!Clever, clean
Fall 2
201
2013
013
01
3
PSYCH / CS 6750
10
Reflective
!Status, image
Fall 2013
PSYCH / CS 6750
11
Taking Off from Maslow
From Anderson, Seductive

Interaction Design
Fall 2013
PSYCH
PSYC
PS
YCH
YC
H / CS 6750
6
15
Simple (Positive) Example
Playful!!
Fun!!
Pleasurable!!
Fall 2013
PSYCH
PSYC
PS
YCH
YC
H / CS 6750
16
16
Simple (Negative) Example

!Conveying success in red #
!More joyful ways to do it?

From Anderson, Seductive
Interaction Design
Fall 2013
PSYCH / CS 6750
17
Exclusivity Snobbery #
Another kind of emotion ..
Fall 2013
PSYCH
PSYC
PS
YCH
YC
H / CS 6750
675
6
750
75
0
118
8
18
Agent-Based Systems
Genetic Algorithm
Definition (What does coexistence mean?)
The word coexistence can be broken into two parts, co- and -exists.
The prefix co- means together and -exist means to be or to live.
Therefore coexistence means being or living together
Coexistence also has another connotation - peaceful
coexistence.
It means more than being in the same place at the same time,
getting along with each other.
Agent-Based Systems
Genetic Algorithm
Session Overview
1

Kansei Systems
Agent-Based Systems
Multi-Agent Systems
Genetic Algorithm
Agent-Based Systems
Genetic Algorithm

Traditional Design Problem
How can we build a system that produces the correct output
given some input?
Agent-Based Systems
Genetic Algorithm

given some input?
Modern-day Design Problem

Modern-day, many of the systems we need to build in practice
have a reactive flavor, in the sense that they have to maintain
a long-term, ongoing interaction with their environment, they
do not simply compute some function of an input and then
terminate.
Agent-Based Systems
Genetic Algorithm

given some input?

terminate.
The main role of reactive systems is to maintain an interaction
with their environment, and therefore must be described in
terms of their on-going behaviors, such as OS, process control
systems, online banking systems, etc.
Agent-Based Systems
Genetic Algorithm

given some input?

terminate.
A still more complex class of systems is a subset of reactive
systems that we call agent
Agent-Based Systems
Genetic Algorithm

given some input?

terminate.
A still more complex class of systems is a subset of reactive
systems that we call agent
Agent is a reactive system that exhibits some degree of
autonomy in the sense that we delegate some task to it, and
the system itself determines how best to achieve this task
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
What is an Agent (Wooldridge and Jennings)
Definition (Agent)
An agent is a computer system that is situated in some
environment, and that is capable of autonomous action in this
environment in order to meet its design objectives
being situated in an environment
autonomy
proactiveness
reactivity
social ability
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Autonomy
Capability of acting independently, exhibiting control over their
internal state
At the end of oneend, we have computer programs such as
conventional word processors and spreadsheets, which exhibits
little or no autonomy. At the other-end of the autonomy
spectrum, you and us. You are completely autonomous. Where
you can ultimately choose to believe what you want - although
society typically constraints your autonomy in various way.
Proactivness
The ability to exhibit goal-directed behavior. If an agent has
been delegated a particular goal, then we expect the agent to
try to achieve this goal.This is in contrast to Object, where it
remains in passive mode until a method invokes on it.
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Reactiveness
Being reactive means being responsive to changes in the
environment. Implementing a system that achieves an effective
balance between goal-directed and reactive behavior turns out
to be hard
Social ability
It is not the ability of exchanges bytes, it is the ability of
agents to cooperate and coordinate activities with other
agents, in order to accomplish assigned goals. This
communication happens at the knowledge level. That is, we
want agents to be able to communicate their beliefs, goals and
plans to one another.
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Other Characteristics of Agents
Mobility: the ability of an agent to move around an electronic

network
Veracity: an agent will not knowingly communicate false
information
Benevolence: agents do not have conflicting goals, and that
every agent will therefore always try to do what is asked of it
Rationality: agent will act in order to achieve its goals, and
will not act in such a way as to prevent its goals being
achieved at least in so far as its beliefs permit
Learning/adaption: agents improve performance over time
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Environments
Accessible vs. Inaccessible (Observable vs. partially

observable)
An accessible environment is one in which the agent can
obtain complete, accurate, up-to-date information about the
environments state
Deterministic vs. non-deterministic

A deterministic environment is one in which any action has a
single guaranteed effect there is no uncertainty about the
state that will result from performing an action
Episodic vs. non-episodic

In an episodic environment, the performance of an agent is
dependent on a number of discrete episodes, with no link
between the performance of an agent in different scenarios
Single Agent vs. Multi-Agents

Which entities have to be regarded as agents? Are the
competitive and cooperative actions?
Agent-Based Systems
Genetic Algorithm
Agent-Bases Systems
Environments
Static vs. dynamic

A static environment is one that can be assumed to remain
unchanged except by the performance of actions by the agent
A dynamic environment is one that has other processes
operating on it, and which hence changes in ways beyond the
agents control
Discrete vs. continuous

An environment is discrete if there are a fixed, finite number of
actions and percepts in it
A chess game as an example of a discrete environment, and
taxi driving as an example of a continuous one.
Continuous environments have a certain level of mismatch
with computer systems
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Agent-Based Systems
Genetic Algorithm
Structure of Agents
Rational Agents
Basic Structure
Perceive the environment through sensors (Percepts)
Act upon the environment through actuators (Actions)
Act rational with respect to a performance measure (e.g. goal:
money, time, energy, utility)
Rational Behavior depends on

Performance measures (goals)
Precept sequences
Knowledge of the environment
Possible actions
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Examples of Rational Agents
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Reflex Agents
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Learning Agents
Agent-Based Systems
Genetic Algorithm
Agent-Based Systems
Example of Learning Agents
Agent-Based Systems
Genetic Algorithm
Multi-Agent Systems
Definition (Multi-Agent Systems (MAS))
A multi-agent system is a computerized system composed of
multiple interacting agents within an environment. Multi-agent
systems can be used to solve problems that are difficult or
impossible for an individual agent to solve.
Definition (intelligent agents)

An intelligent agent in a society is a rational agent with following
abilities,
Reactivity
Proactivity
Social Ability: To interact (communicate, cooperate,
collaborate) with other agents by using Agent communication
language.
Agent-Based Systems
Genetic Algorithm
Multi-Agent Systems
Attribute of MAS
Agent-Based Systems
Genetic Algorithm
Multi-Agent Systems
MAS Coordination
There are two types of agents in MAS

Benevolent Agents
Self Interested Agents
Agent-Based Systems
Genetic Algorithm
Multi-Agent Systems
Benevolent Agents- Contract Nets
Benevolent Agents (cooperation)

Examples are team of fire-brigades, disaster-rescue team, etc.
Agents are assumed to act truthfully
Agents have been designed to help whenever it asks for, which
is called Cooperative distributed problem solving approach
Agent-Based Systems
Genetic Algorithm
Multi-Agent Systems
Self-Interested Agents
Self-Interested Agents (competition)

Examples are arrival management system for airport with
number of different airlines, reverse auction, e-commerce, etc.
Agents tries to maximize its expected utility
Agents know what their options are and what the payoff will be
Strategic deliberation and decision-making
Agent-based system in buying and selling process:
Agent-Based Systems
Genetic Algorithm
Agents for Work-flow and Business Process Management
Work-flow systems aim to to automate the process of a business, ensuring that different business
tasks are expedited by the appropriate people at the right time, typically ensuring that a particular
document flow is maintained and managed within an organization.
ADEPT is an example for an agent-based business process management systems, a business
organization is modeled as a society of negotiating, service providing agents.
Agent-Based Systems
Genetic Algorithm
Agents for Distributed Sensing

The broad idea is to use multi-agent systems to manage networks of spatially distributed sensors.
The sensors, may, for example, be acoustic sensors on a battlefield, or radars distributed across
some airspace.
Sensors will typically provide partial and frequently conflicting data: different parts of the
environment will have different characteristics with respect to the sound and electromagnetic
sensing spectrum.
Agents in the network should cooperate with another, for example, by exchanging information
about when a vehicle pass from one region to another.
Agent-Based Systems
Genetic Algorithm
Agents for Distributed Sensing

The broad idea is to use multi-agent systems to manage networks of spatially distributed sensors.
The sensors, may, for example, be acoustic sensors on a battlefield, or radars distributed across
some airspace.
Sensors will typically provide partial and frequently conflicting data: different parts of the
environment will have different characteristics with respect to the sound and electromagnetic
sensing spectrum.
Agents in the network should cooperate with another, for example, by exchanging information
about when a vehicle pass from one region to another.
Agents for Electronic commerce

Agents for Human-Computer Interfaces
Computer programs that employ AI in order to provide assistance to a user dealing with a
particular application, i.e. A personal assistant who is collaborating with the user in the same work
environment.
Agent-Based Systems
Genetic Algorithm
Agents for Information Retrieval and Management
An information agent is an agent that has access to at least one and potentially many information
sources.
Agent is able to collate and manipulate information obtained from these sources in order to answer
queries posed by users and other information agents.
Agent-Based Systems
Genetic Algorithm
Session Overview
1

Kansei Systems
Agent-Based Systems
Multi-Agent Systems
Genetic Algorithm
Agent-Based Systems
Genetic Algorithm
Genetic Algorithm (GA)

Darwins Principle of Natural Selection

Genetic Algorithms are search and optimization techniques, based
on Darwins Principle of natural selection
Agent-Based Systems
Genetic Algorithm
Genetic Algorithm (GA)

Darwins Principle of Natural Selection

Genetic Algorithms are search and optimization techniques, based
on Darwins Principle of natural selection
Agent-Based Systems
Genetic Algorithm
Genetic Algorithm
DNA :
Is the prime genetic molecule
carrying all the heredity
information
Chromosome
DNA is associated with proteins
and each DNA and its associated
protein is called a chromosome.
Thus, Chromosome is a compact
form of DNA that fits inside the
cell
DNA packaged into a
chromosome can be transmitted
efficiently to both daughter cells
each time a cell divides.
Agent-Based Systems
Genetic Algorithm
Genetic Algorithm
Chromosome
Different kinds of organisms have
different numbers of chromosomes.
Each parent contributes one
chromosome to each pair,
so children get half of their
chromosomes from their mothers and
half from their fathers.
Gene
A gene is the functional and physical
unit of heredity passed from parent to
offspring.
Genes are pieces of DNA, and most
genes contain the information for
making a specific protein.
Agent-Based Systems
Genetic Algorithm
Vocabulary of GA
Genome
Collection of all
chromosomes
Genetic information is
stored in the chromosomes
Chromosome
Each chromosome is build
of DNA
Collection of Genes
Locus
The position of a gene on
the chromosome.
Trait
Aspect (features) of a
chromosome
Agent-Based Systems
Genetic Algorithm
Genetic Algorithm
Algorithm
Design of Genotype (Encoding) : Determine how solutions of

a given problem are represented as genes.
Initialization : Make a given number of individuals (M) with
appropriate genes at the first generation.
Fitness Evaluation : Calculate the fitness of each individual
with an evaluation function. If there is one that satisfies the
termination condition, it is transformed into Phenotype, then
END.
Selection : Choose the necessary number of individuals for
crossover considering fitness
Crossover : Generate individuals of the next generation by
exchanging genes of chosen pairs
Mutation : Change a pair of genes in the predefined way and
with the predefined mutation probability
go to Fitness Evaluation
Genetic Algorithms
Simple example
Example
Individuals are 10 bits sequence. We let the individuals
evolve to the final goal. Which is a bit sequence such that the
first 5 bits are 0 and the latter 5 are 1.
Fitness of the optimal solution is 10
(1) Encoding
Adopt the bit sequence as the genotype (genotype =
phenotype).
Each bit is a gene.
(2) Initialization
Decide the population size = 6. Each gene (0 or 1) is
decided randomly.
(3) Fitness Evaluation

Fitness is defined by sum of two numbers; the number of
zero in the first 5 bits and the number of one in the latter 5
bits.
Fitness of the individual is = 3 +3
Genetic Algorithms
Simple example (2)
(4) Selection in the first generation
Population
Fitness
Selected Individuals (using Selection Techniques)
Fitness Rank
1st generation and the average fitness is 5.0
Selected individuals in the first generation

Average fitness is 5.33
The individuals with the lowest fitness is discarded and replaced by the copy of the highest one.
(4.1) Select the pairs for the crossover
Pairs: (1, 6) and (2,5) and (3,4)
Genetic Algorithms
Simple example (3)
(5) Crossover
st
1st Generation selected individuals before crossover 1 Generation After crossover
1
3
4
Rank
Fitness
Crossover point is selected randomly
Rank
Fitness
Genetic Algorithms
Simple example (4)
(6) Mutation with probability 0.001
For a selected bit , if 0 1 or else 10
1st Generation Before Mutation
1st Generation After Mutation
Rank

Fitness
Fitness
Number of bits to be mutated = 6 *10 *0.001 = 0.06 No mutation will be applied to the bits
Genetic Algorithms
Simple example (4)
(6) Mutation with probability 0.001
For a selected bit , if 0 1 or else 10
1st Generation Before Mutation
1st Generation After Mutation
Rank

Fitness
Fitness
Number of bits to be mutated = 6 *10 *0.001 = 0.06 No mutation will be applied to the bits
Genetic Algorithms
Simple example (5)
(3) Selection in the second generation
2nd Generation - Population
1
2nd Generation Selected individuals

1
Rank
Fitness
Rank
Fitness
Genetic Algorithms
Simple example (6)
(3) Crossover of the second generation
Fitness
Rank
Rank
Fitness
Genetic Algorithms
Simple example (6)
(3) Crossover of the second generation
Fitness
Rank
Rank
Fitness
Agent-Based Systems
Genetic Algorithm
Session Overview
1

Kansei Systems
Agent-Based Systems
Multi-Agent Systems
Genetic Algorithm
Agent-Based Systems
Genetic Algorithm
Introduction
What is Ubiquitous Computing

the method of enhancing computing use by making many
devices (services) available throughout the physical
environment, but making them effectively invisible to the user
computing everywhere:
many embedded, wearable, hand-held devices communicate
transparently to provide different services to the users
devices mostly have low power and short-range wireless
communication capabilities
devices utilize multiple on-board sensors to gather information
about surrounding environments
Agent-Based Systems
Genetic Algorithm
Challenges and Requirements

Hardware, Applications, User Interfaces
Networking, Mobility, Scalability, Reliability
Interoperability, Resource Discovery, Privacy and Security

New trends in ICT: AI, agents, genetics & ubiquitous computing

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

New trends in ICT: AI, agents, genetics & ubiquitous computing

Uploaded by

Copyright:

Available Formats

Trends and Future Dir.

New trends and future directions of ICT

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 1

Trends and Future Dir. in ICT

New trends and future directions of ICT

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 2

Trends and Future Dir. in ICT

Intelligent and Emotional Computing

Emotional Computing, What and Why?

The field is called Affective Technology

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 3

Trends and Future Dir. in ICT

Today, machine prototypes exist that measure human

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 4

Trends and Future Dir. in ICT

Today, machine prototypes exist that measure human

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 5

Trends and Future Dir. in ICT

Intelligent and Emotional Computing

New trends and future directions of ICT, Slide 6

Trends and Future Dir. in ICT

Definition (Neural Network)

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 7

Trends and Future Dir. in ICT

Definition (Genetic Algorithm)

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 8

Trends and Future Dir. in ICT

Definition (Expert Systems)

New trends and future directions of ICT, Slide 9

Trends and Future Dir. in ICT

Artificial Intelligence Software vs. Conventional Software

Artificial Intelligence Software vs. Conventional Software

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 10

Trends and Future Dir. in ICT

Artificial Intelligence (Pros)

Artificial Intelligence (Cons)

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 11

Trends and Future Dir. in ICT

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 12

Trends and Future Dir. in ICT

What is attractive? It is three level of processing

Affective designs impact on usability?

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 13

Taking Off from Maslow

From Anderson, Seductive

Simple (Positive) Example

Simple (Negative) Example

!More joyful ways to do it?

Trends and Future Dir. in ICT

Subha Fernando, Dr.Eng, M.Eng, B.Sc(Special)Hons.

New trends and future directions of ICT, Slide 14

Trends and Future Dir. in ICT