Title: Chapter 4 Decision Support and Artificial Intelligence Brainpower for Your Business
1Chapter 4Decision Support and Artificial
IntelligenceBrainpower for Your Business
2Making decisions by visualizing information as
maps
- A geographic information system (GIS) allows you
to see information spatially in the form of a
map. - The Ice and Marine Services Branch of the
Meteorological Service of Canada provides
accurate and timely reports on sea ice floes in
Canadian waters. The IMSB depends on integrated
GIS and other information technologies to acquire
and process data from data sources such as
satellites, airborne radars and ice/weather
models.
3Making decisions by visualizing information as
maps
- Do you use Web-based map services to get
directions and find the location of buildings?
If so, why? - In what ways could real estate agents take
advantage of the features of a GIS? - How could GIS software benefit a bank wanting to
determine the optimal placements for ATMs?
4Making decisions by visualizing information as
maps
- Remember the 4Ps
- Product
- Price
- Promotion
- Place
- The where of things
5How decisions are made
- One model includes these four phases of decision
making - Intelligence find or recognize a problem, need,
or opportunity - Design consider possible ways of solving the
problem - Choice weigh the merits and consequence of each
solution and then choose one - Implementation carry out the solution
6How decisions are made
- Another model called satisficing is simply making
a choice even though it may not be the best one. - Can be called the just do it model
7Decision making may not be linear.
8Decision making may not be linear.
http//www.witiger.com/powerpoints/goinginternati
onal/sld009.htm
9Types of Decisions
- A structured decision uses certain inputs and
processes them in a precise way guaranteeing a
correct answer e.g. knowing how much GST to
charge on a bill. - A nonstructured decision involves intuition. No
rules or criteria exist guaranteeing choice of
the right answer e.g. introduction of a new
product line. - A recurring decision happens repeatedly.
- A nonrecurring (ad hoc) decision is made
infrequently.
10Types of Decisions
- A structured decision
- Example what is the cost of materials
- A nonstructured decision
- Example will the government continue to
subsidize the program - A recurring decision
- Using a particular shipping partner
- A nonrecurring (ad hoc) decision
- Caterer for the companys 10th anniversary
11Decision Support Systems
- Decision support system (DSS)
- a highly flexible and interactive system that
is designed to support decision making for a
non-structured problem - Decision makers are provided with specialized
support using IT. They must know what information
they need. They must also know how to use the
results of the analysis done by the DSS.
LO1
12Decision Support Systems
- Decision support system (DSS)
- Typical information that a decision support
application might gather and present would be - Comparative sales figures between one week and
the next - Projected revenue figures based on new product
sales assumptions - The consequences of different decision
alternatives, given past experience in a context
that is described - Eg. Selling 4 for the price of 3, bundling
different services
13The decision makersalliance with the DSS
Page 98
LO1
14Three components of a Decision Support System
- The model management system stores and maintains
the DSS models. - Models represent events, facts or situations.
Businesses use models to represent variables and
the relationships between them. - For example, a bank could use a model to see what
impact various increases to the interest rate
would have on their customers mortgage payments.
LO1
15Components of a Decision Support Systems
- The data management component is both the DSS
database management system and information from - the organization
- external sources and
- users.
-
- The user interface management component consists
of the - user interface. This component is where the user
inputs information, commands and models into the
DSS. -
Page 100
LO1
16Example of how the three DSS components work
together
- A user communicates needs to the DSS using the
user interface management component . For example
the user could specify which models to use. Use
of the models is provided by the model management
component of the DSS. The input for the chosen
model(s) is retrieved using the data management
component.
LO1
17Components of a DSS
LO1
18GEOGRAPHIC INFORMATION SYSTEMS
- A geographic information system (GIS) is a DSS
designed specifically to analyze spatial
information. This spatial information can be
shown on a map. - Businesses use GIS software to analyze
information, generate business intelligence, and
make decisions. - Business geography refers to the use of GIS
software to generate maps showing something of
interest to the company e.g. maps showing the
location of homes for sale.
LO1
19GEOGRAPHIC INFORMATION SYSTEMS
- GPS technology is greatly facilitating the
ability of GIS to provide helpful info
http//www.witiger.com/ecommerce/mcommerceGPS.htm
LO1
20Artificial intelligence (AI)
- Artificial intelligence is the use of machines to
imitate the way humans think and behave. For
example, an insurance company could use AI to
detect fraudulent claims. - There are four major categories of AI.
- expert systems
- neural networks and fuzzy logic
- genetic algorithms
- intelligent agents or agent-based technologies
Page 104
LO1
21Artificial intelligence (AI)
- Prof. John McCarthy , Professor of Computer
Science at Stanford University
http//www-formal.stanford.edu/jmc/whatisai/node2.
html
Prof. McCarthy (retired) was a famous Computing
Science professor at Stanford University and he
was responsible for the coining of the term
"Artificial Intelligence"
http//en.wikipedia.org/wiki/John_McCarthy_(comput
er_scientist)
LO1
22Artificial intelligence (AI)
- Artificial intelligence types according to Prof.
John McCarthy , Professor of Computer Science at
Stanford University
Logical AI The program decides what to do by
inferring that certain actions are appropriate
for achieving its goals Search examine large
numbers of possibilities, e.g. moves in a chess
game Pattern recognition For example, a vision
program may try to match a pattern of eyes and a
nose in a scene in order to find a
face. Representation Facts about the world have
to be represented in some way. Usually languages
of mathematical logic are used Inference For
example, when we hear of a bird, we man infer
that it can fly
http//www-formal.stanford.edu/jmc/whatisai/node2.
html
LO1
23Artificial intelligence (AI)
Artificial intelligence types according to Prof.
John McCarthy , Professor of Computer Science at
Stanford University
http//www-formal.stanford.edu/jmc/whatisai/node2.
html
Common sense knowledge and reasoning This is the
area in which AI is farthest from human-level
Learning from experience The approaches to AI
based on connectionism and neural nets specialize
in that Planning Planning programs start with
general facts about the world (especially facts
about the effects of actions) Epistemology This
is a study of the kinds of knowledge that are
required for solving problems in the
world. Ontology Ontology is the study of the
kinds of things that exist. In AI, the programs
and sentences deal with various kinds of objects
LO1
24Artificial intelligence (AI)
- Artificial intelligence types according to Prof.
John McCarthy , Professor of Computer Science at
Stanford University
http//www-formal.stanford.edu/jmc/whatisai/node2.
html
Heuristics A heuristic is a way of trying to
discover something or an idea imbedded in a
program refers to experience-based techniques for
problem solving a heuristic process may include
running tests and getting results by trial and
error. As more sample data is tested, it becomes
easier to create an efficient algorithm to
process similar types of data Genetic
programming a technique for getting programs to
solve a task by selecting the fittest
LO1
25Artificial intelligence (AI)
- Applications of A.I. according to Prof. John
McCarthy , - Professor of Computer Science at Stanford
University
game playing You can buy machines that can play
master level chess for a few hundred dollars.
There is some AI in them, but they play well
against people mainly through brute force
computation--looking at hundreds of thousands of
positions. To beat a world champion by brute
force and known reliable heuristics requires
being able to look at 200 million positions per
second. speech recognition In the 1990s,
computer speech recognition reached a practical
level for limited purposes. Thus United Airlines
has replaced its keyboard tree for flight
information by a system using speech recognition
of flight numbers and city names. It is quite
convenient. On the other hand, while it is
possible to instruct some computers using speech,
most users have gone back to the keyboard and the
mouse as still more convenient.
http//www-formal.stanford.edu/jmc/whatisai/node3.
html
LO1
26Artificial intelligence (AI)
- Applications of A.I. according to Prof. John
McCarthy , - Professor of Computer Science at Stanford
University
understanding natural language Just getting a
sequence of words into a computer is not enough.
Parsing sentences is not enough either. The
computer has to be provided with an understanding
of the domain the text is about, and this is
presently possible only for very limited domains.
computer vision The world is composed of
three-dimensional objects, but the inputs to the
human eye and computers' TV cameras are two
dimensional. Some useful programs can work
solely in two dimensions, but full computer
vision requires partial three-dimensional
information that is not just a set of
two-dimensional views. At present there are only
limited ways of representing three-dimensional
information directly, and they are not as good as
what humans evidently use.
http//www-formal.stanford.edu/jmc/whatisai/node3.
html
LO1
27Artificial intelligence (AI)
- Applications of A.I. according to Prof. John
McCarthy , - Professor of Computer Science at Stanford
University
expert systems A knowledge engineer''
interviews experts in a certain domain and tries
to embody their knowledge in a computer program
for carrying out some task. One of the first
expert systems was MYCIN in 1974, which diagnosed
bacterial infections of the blood and suggested
treatments. It did better than medical students
or practicing doctors, provided its limitations
were observed. heuristic classification An
example is advising whether to accept a proposed
credit card purchase. Information is available
about the owner of the credit card, his record of
payment and also about the item he is buying and
about the establishment from which he is buying
it (e.g., about whether there have been previous
credit card frauds at this establishment)
http//www-formal.stanford.edu/jmc/whatisai/node3.
html
LO1
28A.I.- Expert Systems
Types of A.I.
- An expert (knowledge-based) system is an
artificial intelligence system that captures
expertise in a certain domain and then applies
reasoning capabilities so that a conclusion can
be reached. - For example, an expert system could be used to
diagnose a medical problem. The system could then
recommend a treatment for the condition. The
expert system is useful because previously
medical specialists provided facts and symptoms
that were input into the expert system.
LO2
29Traffic Light Expert System
Types of A.I.
LO2
30Expert systems can
Types of A.I.
- Handle massive amounts of information
- Reduce errors
- Combine information from many sources
- Improve customer service
- Provide consistency in decision making
- Provide new information
- Reduce time personnel spend on tasks
- Reduce cost
LO2
31Expert systems cannot
Types of A.I.
- Capture expertise if domain experts are unable to
explain how they know what they know - Be used for reasoning processes that are too
- complex
- vague
- imprecise or
- require too many rules
- Use common sense
LO2
32Neural Networks and Fuzzy Logic
Types of A.I.
- A neural network (artificial neural network or
ANN) is an artificial intelligence system that is
capable of finding and differentiating patterns. - For example, bomb detection systems in Canadian
airport use neural networks to sense trace
elements in the air.
LO3
33Neural Networks Can
Types of A.I.
- Learn and adjust to new circumstances on their
own - Participate in massive parallel processing
- Function without complete or well-structured
information - Cope with huge volumes of information with many
dependent variables - Analyze nonlinear relationships
LO3
34Neural Networks
Types of A.I.
- Known as the bottom-up approach to the research
and development of intelligent machines, the
neural network approach seeks to replicate in a
computer the actions and functions of biological
neurons found in the human body. - Neurons are cellular transmitters of information
that work by means of the electrical signals that
pass through one neuron to another. - A neural network is, therefore, a group of
neurons that are connected to each other in
complex structures.
LO3
35Neural Networks fo real? ?
Types of A.I.
- Two issues are largely responsible for hindering
full-scale development of artificial neural
networks. - Firstly, the construction of neuron simulators is
cost-prohibitive. - Secondly, current computer architecture still
needs more pathways between components.
LO3
36Fuzzy Logic
Types of A.I.
- Fuzzy logic is a mathematical method of handling
imprecise or subjective information. It assign
values between 0 and 1 to vague or ambiguous
information. Rules and processes, called
algorithms are constructed. These fuzzy logic
algorithms describe the interdependence among
variables. - For example, fuzzy logic is used by Googles
search engine to make sense of the search
criteria that was entered.
LO3
37Fuzzy Logic
Types of A.I.
- handling imprecise or subjective information
http//www.iscid.org/encyclopedia/Fuzzy_Logic
LO3
38Fuzzy Logic
Types of A.I.
- Fuzzy logic expands traditional Boolean or
classical logic in order to allow for partial
truths. - Classical logic requires that a concept be deemed
either true or false, yes or no, black or white
no allowances for the possibility that the answer
may lie somewhere in the middle. - Fuzzy logic, on the other hand, is a superset
that has been developed to manage the gray areas.
http//www.iscid.org/encyclopedia/Fuzzy_Logic
LO3
39Fuzzy Logic
Types of A.I.
- Applications
- Fuzzy logic normally follows the if/then rules
of action and reaction. - For example, if a temperature reaches the desired
setting, then the thermostat switches itself off.
- Basic applications of fuzzy logic can be found in
a growing number of household appliances such as
air conditioners, refrigerators, washing
machines, security systems, etc.
http//www.iscid.org/encyclopedia/Fuzzy_Logic
LO3
40Genetic Algorithms
Types of A.I.
- A genetic algorithm is an artificial intelligence
system that tries to find the combination of
inputs that will produce the best solution.
- Genetic algorithms use
- selection (preference given to better outcomes)
- crossover (portions of good outcomes are combined
in the hope of creating an even better outcome) - mutation (randomly try new combinations
evaluating each combination)
LO4
41Genetic Algorithms Can
Types of A.I.
- Take thousands or even millions of possible
solutions, combine and recombine them until it
finds the optimal solution - Work in environments even if there is no existing
model for finding the right solution
LO4
42Intelligent Agents
Types of A.I.
- Intelligent agent software that assists you, or
acts on your behalf, when performing repetitive,
computer-related tasks - There are four types of intelligent agents
- Information agents
- Monitoring-and-surveillance agents
- Data-mining agents
- User or personal agents
LO5
43Information Agents
Types of A.I.
- Information agents are intelligent agents that
search for information of some kind and return it
to the user. - An example is a buyer agent or shopping bot
which can help a customer find products or
services. When purchasing a book on Amazon.com, a
shopping bot displays a list of similar books
the customer may be tempted to buy.
LO5
44Monitoring-and-Surveillance Agents
Types of A.I. Types of Intelligent Agents
- Monitoring-and-surveillance (predictive) agents
constantly observe and report things of interest.
- For a computer network, a monitoring-and-surveilla
nce agent could be used to look for patterns of
activity and identify potential problems. Agents
could also be used to watch certain Internet
sites looking for stock manipulation or insider
training.
LO5
45Data-Mining Agents
Types of A.I. Types of Intelligent Agents
- A data-mining agent is used to discover
information in a data warehouse. It must sift
through a lot of information. - A common data-mining agent looks for patterns in
information and categorizes items into classes.
For example, a data-mining agent could be used to
find investment opportunities in financial
markets.
LO5
46User Agents
Types of A.I. Types of Intelligent Agents
- User or personal agents are intelligent agents
that take action on your behalf. - For example, a personal agent could assemble
customized news reports to send you. Another
example is Movex software which searches the
Internet negotiating and making deals with
suppliers and distributors.
LO5
47Multi-agent Systems and Agent-based Modelling
Types of A.I. Types of Intelligent Agents
- By observing parts of the ecosystem, artificial
intelligence scientists use hardware and software
models to adapt the ecosystems characteristics
to human and organizational situations. This is
called biomimicry. - For example, biomimicry could be used to predict
how people will behave under certain
circumstances.
LO5
48Agent-Based Modelling
Types of A.I. Types of Intelligent Agents
- Agent-based modelling a way of simulating human
organizations using many intelligent agents, each
of which follows simple rules and adapts to
changing conditions - Multi-agent system groups of intelligent agents
that can to work independently or interact with
each other - Air Canada uses agent-based modelling to find the
optimal route to send air cargo.
LO5
49Swarm Intelligence
Types of A.I. Types of Intelligent Agents
- When individuals in a system consistently follow
a set of rules, complex collective behaviour may
result. - Swarm (collective) intelligence is the
collective behavior of groups of simple agents
that can devise solutions to problems as they
come up and eventually develop a coherent global
pattern. - Swarm intelligence can create and maintain
systems that are flexible, robust, decentralized
and self-organized.
LO5