CSD RQF Artificial Intelligence
Subject: HND in IT
RQF Artificial Intelligence Unit-27
Submitted to: Sabrina Hossain Mou (SHM) Lecturer
School of IT
Designation school: School of IT (BAC International Study
Centre)
Semester: Spring Semester 2021-1
Submitted by: Ehsanul Hoque
Student id: 2019341002
Submission Date: 15/02/2021
Assignment-A1
1
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Table of Contents
Page No
Introduction…………………………………………………………………………….... 3
Investigate the top-down approach of AI and its techniques, and show
how these techniques have been used to build intelligent systems……………………… 4-5
Investigate the bottom-up approach of AI and its techniques, and show
how these techniques have been used to build intelligent systems……………………… 6-7
Develop an intelligent system using a top-down approach with a suitable
programming language or tool…………………………………………………………... 8-12
Test the system and analyse the results against expected results to
identify consistencies………………………………………………………………….... 13-15
Reference……………………………………………………………………………… 16
2
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Introduction: As of late I have begun as an understudy of a reputated Artificial insight
organization named Crafted Mind after finishing Bachelors in IT and I am here to learn under one
of the senior specialized officials of the association Mr. Bayezid Bostami. At a brief timeframe he
requested that I learn and study the fundamentals of computerized reasoning. The seniors have
given me the rules that is considered in making the report which are in the accompanying 1) About
hypothetical establishment of man-made consciousness, latest things and issues everybody should
have clear arrangement, 2) Amateurs are relied upon to comprehend the scope of AI advancements
arising that decide changes in innovation and 3) Should have the option to get information by
student about execution of insight frameworks of various methodologies. Accordingly, I have
created the accompanying report that is helpful for the association.
3
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Investigating the top-down approach of AI and its techniques, and show how these
techniques have been used to build intelligent systems are as follows:
Top-down approach of AI: On the conviction that rationale can be gathered the top-down
methodology is pivoted from a current savvy framework. Stunning outcomes has been appeared
as semantic comprehension in the top-down methodology. Reproducing the rationale behind
language to make discourse and language understanding frameworks is one regular application.
Top-down methodologies are indicating the most outcomes by a long shot when taking a gander
at the fate of solid computerized reasoning. (Artificial Intelligence - Top-down vs Bottom-up Approaches
to Strong AI, 2021)
Techniques of top-down approach: As a reason for dynamic top-down investigation for the most
part alludes to utilizing far reaching factors. Looks to distinguish the higher perspective and the
entirety of its parts is the top-down methodology. Main impetus for the ultimate objective is
typically these parts. Normally connected with "large scale" or macroeconomics is the top-down.
Greatest elements influencing the economy overall is macroeconomics itself in a territory of
financial aspects. Things like the joblessness rates, government finances rate, worldwide and
nation explicit GDP, and expansion rates are remembered for these components. (Top-Down vs.
Bottom-Up: What's the Difference? 2021)
Explaining the top-down approach techniques: Parting configuration approach is trailed by
top-down methodology where configuration follows which breaks a mind-boggling calculation or
an issue into more modest portions called modules this cycle is additionally called as
modularization.
Steps followed in top-down design approach:
❖ Dividing it into at least two sections by taking the mind-boggling entire issue.
❖ Split them further whenever split parts end up being too huge to be tackled.
❖ According to the sub-issue progression combine arrangements.
❖ Whole issue is tackled if all parts have been effectively settled. (Singh, 2021)
4
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Natural language processing: I work 40 hours per week for us to be this poor. Identifying the
mockery in the above statement most people would know regardless of whether it takes them a
second or two. The estimation communicated in the above sentence however envision causing a
PC to comprehend it as such a test being handled. For slant investigation they are managing the
essential issues of normal language preparing. The language we use for speaking with one another
as common language is somewhat not quite the same as the manner in which we speak with PCs.
Characteristic language is mind boggling, equivocal, turbulent. Programming dialects, for
example, built dialects, hold fast to severe principles and rationale. (Bhunia, 2021)
Knowledge based system: Automated frameworks that copy human thinking are information-
based frameworks. In specific areas of utilization such frameworks are worked with explicit
information and work in a manner like that of a human master. For the most part undeniable level
human knowledge in emblematic structure they were made trying to catch and copy ordinarily in
a bunch of on the off chance that, at that point rules. (Knowledge Based Systems - an overview |
ScienceDirect Topics, 2021)
Fuzzy logic: The idea can have a level of truth anyplace somewhere in the range of 0.0 and 1.0
fluffy rationale is a speculation of standard rationale. Applies just to ideas that are totally obvious
in standard rationale or totally bogus. About intrinsically unclear ideas, for example, 'stature fluffy
rationale should be utilized for thinking. (American, 2021)
An example of using top-down technique used to build an intelligent system is:
A wonder known as the Stroop impact is one exemplary illustration of top-down handling in real
life. Individuals are demonstrated top notch of words imprinted in various tones in this assignment.
As opposed to the actual word they're at that point requested to name the ink tone. At the point
when the significance of the word and the ink tone doesn't coordinate curiously, individuals are
much increasingly slow more mix-ups. Individuals make some harder memories when "red" is
imprinted in green ink rather than red ink for instance. (Top-Down Processing: How It Influences
Perception, 2021)
5
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Investigating the bottom-up approach of AI and its techniques, and show how these
techniques have been used to build intelligent systems is listed below:
Bottom-up approach: Here are however, some bottom-up approaches resembling the creation
of child machines with a disposition to be told that are value noting the bottom-up approach has
produced genetic algorithms which are created to assist notice optimal solutions to complicated
problems. the foremost ideal thanks to produce human-like intelligence as we tend to ourselves is
bottom-up approach would be a part of a bottom-up design process. (Artificial Intelligence - Top-down
vs Bottom-up Approaches to Strong AI, 2021)
Techniques of bottom-up approach: Totally extraordinary methodology is the base up
examination. Base up methodology by and large zeros in its examination on explicit qualities and
miniature credits of an individual stock. Putting focus in base up is on area-by-area essentials.
Beneficial chances through the mannerisms of an organization's credits and its valuations are
utilized in investigation that tries to distinguish in contrast with the market. At the organization
level yet doesn't stop there is base up contributing starts its exploration. (Top-Down vs. Bottom-Up:
What's the Difference? 2021)
Explaining the bottom-up approach techniques: Blending configuration approach is adopted
by base up strategy which consolidates the more modest sub-module into the bigger complex
module. From basic parts associated models are fabricated.
In Bottom-Up design approach following steps:
❖ Sub-some portion of the perplexing module takes the littlest.
❖ Sub part to be tackled.
❖ Iteratively to get the last arrangement combine all the littlest sub-part. (Singh, 2021)
Artificial Neural Networks: On the conviction that working of human cerebrum is the possibility
of fake neural organizations is based by making the correct associations, can be imitated utilizing
silicon and wires as living neurons and dendrites. Made out of 86 billion nerve cells called neurons
is the human mind. By axons they are associated with other thousand cells. Contributions from
tangible organs are acknowledged by dendrites or Stimuli from outer climate. Going through the
neural organization are these sources of info that make electric driving forces rapidly. Deals with
the issue or doesn't send it forward a neuron would then be able to send the message to another
neuron.
6
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Types of Artificial Neural Networks (ANN)
Feed Forward (ANN): The data stream is unidirectional in this counterfeit neural organization.
To other unit from which it doesn't get any data a unit sends data where there are no criticism
circles. In example age they are utilized. They have fixed data sources and yields.
Feedback (ANN): Criticism circles are permitted. They are utilized in substance addressable
recollections. (Artificial Intelligence - Neural Networks - Tutorialspoint, 2021)
Evolutionary computing: Ordinarily includes the utilization of some programmed technique for
creating and assessing progressive ages of a program in developmental registering until a
profoundly capable arrangement advance. (Artificial intelligence - Evolutionary computing, 2021)
Swarm intelligence: To be a complete strategy to tackle complex advancement issues the
multitude knowledge calculations have been demonstrated by mimicking the development
practices of natural multitudes. Information science is standing out enough to be noticed, these
days which needs snappy administration and examination of enormous information. To consistent
and differentiable capacities most conventional strategies must be applied. (Yang et al., 2021)
An example of using bottom-up technique used to build an intelligent system is:
A base up method in which at least two atoms respond in compound fume statement the fume stage
inside a synthetic fume testimony reactor prompting arrangement of nanomaterial over a substrate.
Arrangement of the ideal material or the union can be liked by controlling the pressing factor and
temperature conditions. (2021)
7
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Developing an intelligent system using a top-down approach with a suitable programming
language or tool is given below:
Here I have created a game application of Tic Tac Toe when first opening the application the user
gets options to select different game modes later.
Figure-1
After selecting Tic Tac Toe play it lets the player choose as there is modes single player and
multiplayer. When single player is selected it lets player play with the programming of the
application.
Figure-2
After selecting single player it lets us select a level and then we have to select it.
8
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Figure-3
Before starting the game for the first time it tells us how to play the game just for the first time.
Figure-4
Then it shows a tutorial on how to play the game and wher to place it. It tells us to place the symbol
at the centre.
9
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Figure-5
Then the application program tells us to place symbol at the bottom left side.
Figure-6
Then it indicates to place another symbol diagonally on upper right side.
10
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Figure-7
After that it shows that the symbols are connected in a line meeting the goal.
Figure-8
11
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
After winning the game it shows victory and score one and there is three options for the player to
select which are home screen, Play and leader board.
Figure-9
12
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Testing the system and analysing the results against expected results to identify consistencies
are as follows:
Here in the second level I have started testing the system of the tic tac toe application programming.
Figure-10
Here I have placed symbol in the centre and the opponent has place symbol on the left side in
Figure-11.
Figure-11
13
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
As player I have placed symbol on the bottom left and the opponent has placed symbol to the top
right side to stop me from winning in the next move in Figure-12.
Figure-12
In Figure-13 I have placed symbol on the bottom right side and opponent has placed symbol on
the centre top and did not stop me from winning in level 2.
Figure-13
14
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
At Figure-14 I have made the three cross and did not expect the opponent to let me win.
Figure-14
Here in Figure-15 I have cleared level 2 of tic tac toe in the game.
Figure-15
15
Prepared by Ehsanul Hoque
CSD RQF Artificial Intelligence
Reference
2021. [online] Available at: <https://www.quora.com/What-is-the-bottom-up-technique-in-nanotechnology>
[Accessed 11 February 2021].
2021. [online] Available at: <https://www.infosys.com/insights/ai-automation/ambient-intelligence.html> [Accessed
9 March 2021].
American, S., 2021. What is 'fuzzy logic'? Are there computers thatare inherently fuzzy and do not apply the usual
binary logic?. [online] Scientific American. Available at: <https://www.scientificamerican.com/article/what-is-fuzzy-
logic-are-t/> [Accessed 11 February 2021].
Bhunia, P., 2021. EXCLUSIVE – Coupling Top-Down And Bottom-Up Approaches To Natural Language Processing.
[online] OpenGov Asia. Available at: <https://opengovasia.com/exclusive-coupling-top-down-and-bottom-up-
approaches-to-natural-language-processing/> [Accessed 16 January 2021].
Existek.com. 2021. [online] Available at: <https://existek.com/blog/ai-programming-and-ai-programming-
languages/> [Accessed 5 February 2021].
Encyclopedia Britannica. 2021. Artificial intelligence - Evolutionary computing. [online] Available at:
<https://www.britannica.com/technology/artificial-intelligence/Evolutionary-computing> [Accessed 11 February
2021].
Flickspin.com. 2021. Artificial Intelligence - Top-Down Vs Bottom-Up Approaches To Strong AI. [online] Available
at: <http://www.flickspin.com/en/artificial_intelligence/ai_top_down_vs_bottom_up> [Accessed 15 January 2021].
Investopedia. 2021. Top-Down Vs. Bottom-Up: What's The Difference?. [online] Available at:
<https://www.investopedia.com/articles/investing/030116/topdown-vs-bottomup.asp> [Accessed 15 January 2021].
Sciencedirect.com. 2021. Knowledge Based Systems - an overview | ScienceDirect Topics. [online] Available at:
<https://www.sciencedirect.com/topics/computer-science/knowledge-based-systems> [Accessed 11 February 2021].
Singh, J., 2021. Top-Down And Bottom-Up Design Approach - Embhack. [online] EmbHack. Available at:
<https://www.embhack.com/top-down-and-bottom-up-design-approach/> [Accessed 16 January 2021].
Tutorialspoint.com. 2021. Artificial Intelligence - Neural Networks - Tutorialspoint. [online] Available at:
<https://www.tutorialspoint.com/artificial_intelligence/artificial_intelligence_neural_networks.htm> [Accessed 31
January 2021].
Verywell Mind. 2021. Top-Down Processing: How It Influences Perception. [online] Available at:
<https://www.verywellmind.com/what-is-top-down-processing-2795975> [Accessed 11 February 2021].
Yang, J., Qu, L., Shen, Y., Shi, Y., Cheng, S., Zhao, J. and Shen, X., 2021. Swarm Intelligence in Data Science:
Applications, Opportunities and Challenges.
16
Prepared by Ehsanul Hoque
