Engineering Techniques

randombroadAI and Robotics

Oct 15, 2013 (4 years and 26 days ago)

125 views

Engineering Techniques


Machine Learning

Machine learning is an area of artificial intelligence concerned with the study of
computer algorithms that improve automatically through experience. In practice, this
involves creating programs that optimize a perf
ormance criterion through the analysis of
data. Among the many techniques about machine learning, review was done on Model
selection and Decision trees


Decision Trees

Good for Prediction and making inferences but only when a training
set or

a learning se
t
is available.

The dataset must have instances for which the actual
value of the target
variable is

known and the associated predictor variables.


Neural Networks


Game theory

Game theory is the formal study of conflict and cooperation. Game theoretic con
cepts

apply whenever the actions of several agents are interdependent. These agents may be

individuals, groups, firms, or any combination of these. The concepts of game theory

provide a language to formulate, structure, analyze, and understand strategic sc
enarios.

The internal consistency and mathematical foundations of game theory make it a prime

tool for modeling and designing automated decision
-
making processes in interactive
environments.

As a mathematical tool for the decision
-
maker the strength of gam
e theory is the

methodology it provides for structuring and analyzing problems of strategic choice.

The process of formally modeling a situation as a game requires the decision
-
maker to
enumerate explicitly the players and their strategic options, and to c
onsider their
preferences and reactions. The discipline involved in constructing such a model already
has the potential of providing the decision
-
maker with a clearer and broader view of the
situation. This is a “prescriptive” application of game theory, w
ith the goal of improved
strategic decision making.


It typically involves several
players
; a game with only one player is usually called a
decision problem
. The formal definition lays out the players, their preferences, their
information,
and the

strateg
ic actions available to them, and how these influence the
outcome.




communication between components


Agent technology is well
-
suited for use in applications that reason about the messages or objects received
over a network.


Again, agents are most suite
d to applications that require communications between
components, sensing or monitoring of the environment, or autonomous operation.
Since agents have the ability to reason (i.e. draw inferences), they can easily
perform sequences of complex operations bas
ed on messages they receive, their
own internal beliefs, and their overall goals and objectives.

Strengths and Weakness of Decision Tree Methods

The strengths of decision tree methods are:

o

Decision trees are able to generate understandable rules.

o

Decisio
n trees perform classification without requiring much computation.

o

Decision trees are able to handle both continuous and categorical
variables.

o

Decision trees provide a clear indication of which fields are most
important for prediction or classification.


The weaknesses of decision tree methods

o

Decision trees are less appropriate for estimation tasks where the goal is to
predict the value of a continuous attribute.

o

Decision trees are prone to errors in classification problems with many
class and relativ
ely small number of training examples.

o

Decision tree can be computationally expensive to train. The process of
growing a decision tree is computationally expensive. At each node, each
candidate splitting field must be sorted before its best split can be f
ound.
In some algorithms, combinations of fields are used and a search must be
made for optimal combining weights. Pruning algorithms can also be
expensive since many candidate sub
-
trees must be formed and compared.

o

Decision trees do not treat well non
-
re
ctangular regions. Most decision
-
tree algorithms only examine a single field at a time. This leads to
rectangular classification boxes that may not correspond well with the
actual distribution of records in the decision space.