REVIEW PAPER ON MACHINE LEARNING

PREPARED BY SHUBHAM SAH (18BCS6537)

BIBEK RAWAT (18BCS6537)

Abstract
This paper provides an extensive review of studies related to expert estimation of software
development using Machine-Learning Techniques (MLT). Machine learning in this new era, is
demonstrating the promise of producing consistently accurate estimates. Machine learning
system effectively “learns” how to estimate from training set of completed projects. The main
goal and contribution of the review is to support the research on expert estimation, i.e. to ease
other researchers for relevant expert estimation studies using machine-learning techniques.
This paper presents the most commonly used machine learning techniques such as neural
networks, case based reasoning, classification and regression trees, rule induction, genetic
algorithm & genetic programming for expert estimation in the field of software development.
In each of our study we found that the results of various machine-learning techniques depends
on application areas on which they are applied. Our review of study not only suggests that
these techniques are competitive with traditional estimators on one data set, but also illustrate
that these methods are sensitive to the data on which they are trained.
Introduction

It is a branch of artificial intelligence. AI is mainly use for making a machine that is able to
work like human. So machine learning is quite similar to artificial intelligence. But there is big
difference between machine learning and artificial intelligence. Machine learning learn itself
like its own program and it make that program far better and useful than previous one then
it is cable to perform the task in real world. So we can that it is like data mining. Machine
learning is all about the data. Nowadays its very necessary because it first analyse the data
then response the given instruction and at last it give us real feel like real man.it mainly focus
in similar pattern then it makes clusters. And clusters is a collection of similar type of pattern.
This clusters can analyse the data and processing the given data.
At a general level, there are two types of machine learning: inductive, and deductive. Deductive
learning works on existing facts and knowledge and deduces new knowledge from the old.
Inductive machine learning methods create computer programs by extracting rules and
patterns out of massive data sets. Inductive learning takes examples and generalizes rather
than starting with existing knowledge one major subclass of inductive learning is concept
learning. This takes examples of a concept and tries to build a general description of the
concept. Very often, the examples are described using attribute-value pairs.

Machine learning overlaps heavily with statistics. In fact, many machine-learning algorithms
have been found to have direct counterparts with statistics. For example, boosting is now
widely thought to be a form of stage wise regression using a specific type of loss function.
Machine learning has a wide spectrum of applications including natural language processing,
search engines, medical diagnosis, bioinformatics and cheminformatics, detecting credit card
fraud, stock market analysis, classifying DNA sequences, speech and handwriting recognition,
object recognition in computer vision, game playing and robot locomotion.

There are some key terminology in machine learning.

1. Data and information
It can convert data into any kind of information. Lets us consider an example of ice-cream.it
has cone, cream and cold in test that is our data. And the ice-cream is our information.
2. Problem solving tool
We have different type of problem like puzzle problem, sales problem. All the time we cant
solve them machine should be
able to solve these kind of problem. And it solve by its own algorithm. Let us consider an
example of chess game a simple machine can win our mind that is only possible of machine
learning.
3. Combination of computer science, engineering and statistics.
Computer science refers the different kind of programming language, engineering refers how
to apply the logic in the
real world and statistics refers the all the kind of mathematical operation that is used in
machine learning.
4. Interprets data and act on it
That means it takes an input from the user and perform the task according to user command.
 5. Optimise performance criteria using past experience.
Its brain depend on its work experience. Lets us take an example of sales manager if his work
experience is long. He/she can take good decision for their company. There are three types
of machine learning and they are as follows:

 Supervised learning
It is a teacher oriented learning.in this learning system already know the what is the output
of the given instruction.
It is similar like parents teaches his son a for apple and b for ball.

 Unsupervised learning
Its is just opposite to the supervised learning .it don’t have an teacher who can teach.so we
can not predict the
output of our command.so system recognize the pattern firstly then it perform its task.

 Reinforcement learning
It is a reward base learning and feedback oriented learning. Here we can get positive and
negative feedback if
the feedback is positive then our work is success else our work is unsuccess. We can get
feedback from our working
environment and reward can get by programmers code.
Steps in machine learning:
 collect the data.
 prepare the input data.
 analysing the input data.
 developing the algorithm.
  test the algorithm.
 use it.
First we have to collect data from different sources that can be images, text or program also
then after collecting the data we should manage this data in outputable (executable) form
then the we should analysing the our processed data and the next step is developing the
algorithm. After developing the algorithm we have to test it ,it is working properly or not. And
our last and final step is its working process in real world.so these are the step we have to
follow for any new machine learning process.

Conclusion:
The main contribution of this review is to discuss the various Machine-Learning
Techniques employed in effort estimation, cost estimation, size estimation and other field of
Software Engineering. Machine learning is concerned with computer programs that improve
their performance through their experiences with their audiences. Machine learning
improves the technological world with its features. The main objective of the machine
learning is to make the machines or devices learn the human behaviours and read their minds
by studying their regular activities. It is the future of the modern society as it allows the
devices to act as human. In the present world, the reason behind the development of the
country is the way they implement machine learning in different aspects. Learning means
improving with experience at some task.
Nowadays, in modern society every country follow the machine learning concept. By
following it, they developed modern technologies like facial reorganization, surveillance
system, recommendation engines, optical character reorganization, advertisements and
business ideas. The learning process involves with the real world first. Then the measuring
devices act which senses the environment and gives the input which further get processed
and then their dimensions are reduced which leads to the model learning and this concludes
with the testing and based on its data analysis, the final product is prepared.

REFERENCES
1. www.googlescholar.com
2. www.geeksforgeeks.com
3. www.edx.com
4. https://www.expertsystem.com/machine-learning-definition/
5. https://en.wikipedia.org/wiki/Machine_learning#Supervised_and_semi-
supervised_learning
6. https://www.mathworks.com/discovery/machine-learning.html