Analysis of Unsupervised Based Techniques for Artificial Intelligence

Desklib provides past papers and solved assignments for students. This report analyzes unsupervised AI techniques.

UNSUPERVISED BASED
TECHNIQUES FOR AI

Contents
Introduction......................................................................................................................................3
Outline.............................................................................................................................................5
A: Outline of methods, techniques, and theory used in AI..........................................................5
B: Different techniques used in AI and their evolution...............................................................5
Analysis...........................................................................................................................................8
Comparison of different AI techniques.......................................................................................8
Review.............................................................................................................................................9
Conclusion.....................................................................................................................................11
References......................................................................................................................................12

List of figures
Figure 1: Artificial Intelligence.......................................................................................................5
Figure 2: Reasoning based on cases................................................................................................7
Figure 3: AI techniques...................................................................................................................8
List of table
Table 1: Table for comparison of different AI techniques..............................................................9

Introduction
AI stands for “Artificial Intelligence” is a technique used for the development of different digital
devices and applications such as computer, robot, etc. with smart thinking same like a human
being to enhance multiple functions of devices and applications especially related with the
knowledge of a human being. It consists of coding, computation, and statics for the processing,
calculation, and working of the AI system (Selvamanikkam, 2018). It consists of different
intelligence capability which is given below:
 Problem-solving
 Reasoning
 Perception
 Learning and speaking ability
 Smart thinking
Traces of the AI system can be found in the history which was used for problem-solving.
Multiple research and developments can be recorded in the field on AI from the last few decades
for the better understanding and enhancement of the AI system in each and every field of our day
to day life.
Nowadays, it is most commonly used in our devices, business, and industries to solve human-
related works and problems in an easy way along with a quick response. It made revolutionary
changes in our lives and reduce the efforts to solve complex works and problems.

Figure 1: Artificial Intelligence
(Source: Mirman, 2018)

Outline
A: Outline of methods, techniques, and theory used in AI
In the AI system, multiple techniques and algorithms are used for the development of smart
thinking capacity for the system for efficient problem solving and efficient computation of
complex problems. It consists of mathematics such as linear algebra, probability, and their
functions along with optimization of multiple works (Dezhic, 2018). Different techniques and
algorithms are used for the development of AI. Some of them are given below:
 Classical programming
 Machine learning
 Knowledge-based system
 Smart search
 An algorithm based on genetic problems
 Knowledge of probability and graphical models
All these methods, techniques and theory used for effective design and development of a system
with smart intelligence to work, think and behave the same like a human being such as computer
system, robots, and others.
B: Different techniques used in AI and their evolution
Various techniques are developed by experts and developers to design better and improved AI
system to simplify our day to day life. Some of the famous techniques are given below:
Reasoning based on cases: It is also known as CRB (Case-based reasoning) used to
identification of simple steps to solve any problem by repetition of multiple attempts. It consists
of four steps such as retrieval of a recent case similar to the problem, solution of the problem on
the basis of the retrieved case, revision of solution and storage of solution along with simulation
for the future use (Chen et al. 2008).

Figure 2: Reasoning based on cases
(Source: Chen et al. 2008)
Ruled based system: It is also known as RBS which works on a set of rules already defined by
the experts and professionals which are available in the database of the system. It helps in
calculation and simplification of the task with the help of a defined set of rules and identification
of the most relevant rule for the task (Chen et al. 2008).
Neural network: It is also known as an artificial neural network and referred to as “ANN”. It
works the same as the human brain and helps in effective calculation utilizing human thinking
ability of the system. It consists of different subunits such as nodes, neurons and other along with
interconnected network for their efficient working (Chen et al. 2008).
Genetic algorithm: It works on the algorithm of genetic formula based on the calculation and
selection of neutral. In this technique, given task or problem is solved repeatedly until most
significant results are recorded and the process is saved for the future use and reference (Chen et
al. 2008).
Cellular automata: It consists of a cellular structure for the simplification of a problem or task
using a discrete system of time, state and space and works on the basis of the interaction between

neighbor cells. It is further classified into four types and classes depending on the working of the
system (Chen et al. 2008).
Figure 3: AI techniques
(Source: Razor Robotics, 2018)

Analysis
Comparison of different AI techniques
With the development of new techniques and technologies by the experts and developers resulted
in significant changes and improvements in the Artificial Intelligence system and their
performance due to which power quality is improved much better in comparison with the system
with old methods and techniques (Bhaskar et al. 2016). A comparison table on the basis of the
power quality of the AI system is given below:
Table 1: Table for comparison of different AI techniques
SRF theory Fuzzy logic controller Particle swarm
optimization
PI controller
It consists of three
phase system
converted into two
phases for the
simplification and
minimization of
complex structure and
problems.
It is also known as FLC
which is a very
excellent method used
in the application of the
control system.
In this, the processing
of data is done on the
basis of partial
membership.
It is also known
as PSO and
work on the
basis of social
behavior such as
fish schooling.
It helps in
reducing total
harmonic
distortion of a
system to a very
low level.
It consists of
integration and
proportion value
calculated for the
simplification of the
task.
In this technique, two
values of the
calculation are shown
as Ki and Kp for the
output.
(Bhaskar et al.
2016)

Review
For the better understanding of Artificial Intelligence and their techniques, a review over various
papers and journals has been done and their details have been explained below:
Artificial intelligence in medicine: Effective design of an AI system to eliminate the challenges
related to the diseases and medical filed has resulted in better management of medical clinics,
payment methods, and others. In this paper, two important approaches named “Dynamic
decision and Markov decision” has been used for the implementation of the AI system in the
medical field. It resulted in an effective technique to reduce the issues related to the medical field
and treatment of diseases (Bennett and Hauser, 2013).
Research Priorities for Robust and Beneficial Artificial Intelligence: In this paper, the
development of a robust and beneficial AI system has been explained to enhance the working
efficiency and to reduce the failure of the system. Multiple benefits have been explained with
various examples for the effective analysis of AI system and methods to improve the working
efficiency in a much better way for the benefit of human being, business, companies, etc. due
which various companies are investing large amount for the development of robust and more
beneficial AI (Russell et al. 2015).
Probabilistic machine learning and artificial intelligence: In this paper, probabilistic
framework has been used as one of the technique for the development of AI system, robots, etc.
which helps in high analysis of uncertain data and predictions in the various field to improve the
advance planning and effective decision system in business and industries (Ghahramani, 2015).
A Panorama of Artificial and Computational Intelligence in Games: In this paper, high
efficiency of AI system in the field of computer and mobile games has been shown on the basis
of ten major pieces of research done in the same field such as player modeling, computational
narration, searching and planning system, etc. The main criteria for the analysis of different
techniques were dominant methods of AI implementation, interactive system for the
communication between players and relationship of the game with the human being. All these
techniques used in the paper has improved knowledge and major factors influencing the AI
system in the field of computer games to help developers and experts for future researches
(Yannakakis and Togelius, 2015).

Artificial intelligence in healthcare: past, present, and future
In this paper, the implementation of the AI system in the field of healthcare has been explained
helping in the rapid growth of treatment and diagnosis facilities for healthcare. In this paper,
different techniques such as Neural network, classical SVM, deep learning and neutral method
has been used for the improvement and development of effective AI tools for various diseases
especially for Cardiology and cancer treatments (Jiang et al. 2017).

Conclusion
In this report, Artificial intelligence system has been discussed along with their significant
benefits and improvements in various fields to maximize work efficiency and minimize the
complexity of various works. Different techniques, methods, and theories used in the AI system
have been discussed to improve the knowledge and awareness about the AI system and its
implementation in different sectors to reduce the human effort. Also, the review of different
research on the AI technique improved deep knowledge about the role and techniques used in the
different application of the AI system as well as devices.