Filetype PDF | Posted on 27 Jan 2023 | 3 years ago
Authentication
digital resources
218x Filetype PDF File size 0.48 MB Source: rspsciencehub.com
www.rspsciencehub.com Volume 02 Issue 07 July 2020
Machine learning amalgamation of Mathematics, Statistics and Electronics
1 2 3 4
Trupti S. Gaikwad , Snehal A. Jadhav , Ruta R. Vaidya , Snehal H. Kulkarni
1,2,3,4 Trupti S. Gaikwad, Dept. of Computer Science, Vishwakarma College of Arts, Commerce and
Science, Maharashtra, India
1
trupti08jadhav@gmail.com
Abstract
Interdisciplinary research is a manner of research carried out by an individual or group of persons. The
knowledge, data, techniques, concepts are incorporated from two or more disciplines. In this paper we
tried to throw light on this concept. Machine learning is a branch of computer science which uses the
information, tools for collection of data, methods for analysis from the subjects like Electronics,
Mathematics and Statistics. Why we use machine learning? Because it plays an influential role in
prediction of data. Machine learning is used to find hidden patterns and essential ideas from data as
well as it solve complex problems. In today’s world, many applications have large volume of data like
structured, unstructured and semi structured. This unexploited resource of knowledge can be used to
improve business decisions. As data diversifies many are adapting to machine learning tool for analysis
of data, so that, they can exploit intelligence and benefit from that data at most. Machine learning
adopts different algorithms and each algorithm performs different functionality. In this paper, we tried
to explain through example, how Electronics is used for collection of data while Mathematics and
Statistics are used for analysis and finally using Machine learning results can be predicted.
Keywords: Machine learning, Analysis, Statistics, Electronics, Mathematics
science. In conventional programming, algorithms
1. Introduction are explicitly written for problem solving used by
Machine learning is intersection of Mathematics,
Statistics, probabilistic, Electronics, and Computer computer. On other hand machine learning is a
Science. Machine learning uses algorithm to learn concept that a machine can learn from past
iteratively from data and using positive feedback experience and previously solved examples,
can build an intelligent application. Though without being explicitly programmed. For
machine learning differs from traditional example, biometric attendance, fraud detection,
computational access, it is part of computer face recognition, text recognition etc.
International Research Journal on Advanced Science Hub (IRJASH) 100
www.rspsciencehub.com Volume 02 Issue 07 July 2020
1.1 What is Machine Learning? estimate the mapping function so that for new
Machine learning is the study of computer input data (X), the algorithm can predict the
algorithms that improve automatically through output variable (Y) for that data.
experience [1-4]. It provides ability to system, The process of learning of the algorithm
learn and improve using previous experience from the training data set, it can be
without being explicitly programmed. It mainly considered as a teacher supervising the
focuses on the development of a program so that learning process so that it is called
they can access data and learn it for themselves. supervised learning. It can be further
grouped into regression and classification
Electronics problem.
Examples of supervised learning algorithm
are
Maths 1. For classification problems-
Machine Statistics Support vector machines
learning
2. For regression problem- Linear
regression
Fig.1. Interdisciplinary Machine Learning 3. For classification and regression
How machine learning works? problem- Radom forest.
The learning process starts with observations or Unsupervised machine learning-
data, like direct experience, example or instruction, Unsupervised machine learning has only
so that it will find patterns for data and make better input data (X). It does not have related
determination about future based examples output variable (Y). To acquire more
provided. The main goal is to allow the system to knowledge about the data underlying
learn automatically without human intervention structure is designed or distribution in the
and modify actions accordingly. It has different data is done. It is called unsupervised
types of algorithm. Machine learning algorithms learning because it does not have correct
are classified as: answer and teacher. Algorithms itself can
Supervised machine learning- In supervised discover and present the interesting
machine learning input variable (X) use by an structure in the data. It can further group
algorithm to give an output variable (Y). Here into clustering and association problems.
the algorithm learns the mapping function from Unsupervised learning algorithm uses
input to the output. The mapping function is Y following algorithms
= f(X). The main aim of an algorithm is to 1. For clustering problem-k means
International Research Journal on Advanced Science Hub (IRJASH) 101
www.rspsciencehub.com Volume 02 Issue 07 July 2020
2. For association rule mining-
Apriori algorithm
Semi supervised machine learning- Regression
In semi supervised machine learning only Supervised
for some of input variable(X), an output Learning
Classification
variable (Y) is present. It is a composite of
supervised machine learning and Clustering
Machine Unsupervised
unsupervised machine algorithm. Here we Learning Learrning
Algorihm Association
can use both Problems
i) Supervised learning technique where Semi-supervised
Learning
we want best prediction for unlabeled
Reinforcement
data, so that we can feed that data back Learning
to supervised learning algorithm as
training data and use that model to Fig.2. Classification of Machine Learning
Algorithm
make prediction for new unseen data.
ii) Unsupervised learning technique to 1.2 Electronics: Elementary component of
find and learn the structure in the input machine learning
variables.
Most of today's real world problems are Electronics is primary element whenever we talk
from this type as problems can be about automated or intelligent or smart systems.
expensive or cheap. For expensive label Electronics and different branches of computer
problems domain experts are required science like machine learning, artificial intelligence
and cheap unlabeled problems are easy are blended together to invent new applications.
to collect and store data. The use of machine learning in engineering field is
very vital for signal processing. Due to this, there is
Reinforcement Learning- The machine is increase in accuracy and quality when sound,
exposed to environment and it learns itself images, and other inputs are transmitted. Machine
by using positive feedback and negative learning algorithms are helpful to model signals,
feedback. The machine learns from past for pattern detection, to draw meaningful
experience. It is directed to make specific inferences, and make precise adjustments to signal
decisions. It tries to capture the best output.[5-8]
possible knowledge to make accurate To feed the data in machine learning systems signal
business decisions. processing techniques are useful.
International Research Journal on Advanced Science Hub (IRJASH) 102
www.rspsciencehub.com Volume 02 Issue 07 July 2020
Digital signal processing and digital image Singular Value Decomposition (SVD) etc.
processing these two are used in many applications Natural language processing on tabular datasets,
along with machine learning. data files such as encoding and dimensionality
reduction, images etc is applications of Linear
Smart sensors are also used along with machine Algebra in machine learning.
learning for developing number of applications like Multivariate Calculus: The most of machine
weather forecast system, in healthcare instruments, learning algorithms are trained on multiple
in smart home systems etc. variables. It helps to better quantify and predict the
1.3 Mathematics behind machine learning information accurately. Laplacian and Lagragian
Distribution, Vector-Values Functions, Directional
Mathematics is main part of machine learning as it Gradient, Differential and Integral Calculus,
is used at backend. Machine learning acquires data Partial Derivatives, Jacobian are the methods of
through algorithms and then uses this data to make Multivariate Calculus use in machine learning.
predictions. Machine Learning requires Examples of Multivariate Calculus include
mathematical knowledge. It includes linear calculation of monthly rain fall, temperature and
Algebra, Calculus, Statistics, Discrete Maths, wind speed and so on.
Probability and Optimization which help to create Graph theory: Graphs represent flow of
algorithms. It includes accuracy, training time, computation. Graph learning models can be used
model complexity, choosing parameter setting and to learn machine learning algorithms. Graphs
validation strategies. Importance of Math’s topics represent computationally various matrices while
needed for machine learning is: matrix provides different types of information. The
Linear Algebra: It is the backbone of machine machine learning models having graph like
learning. To find the values of variables X and Y structure are K-mean, K-nearest neighbours,
matrix operations are used which are parts of Decision trees, Random forest, neural networks.
linear algebra. Due to this, linear algebra is
necessary in machine learning. Not only all the 1.4 Statistics: Prerequisite of machine learning
operations in Linear Algebra are systematic rules Statistics plays important role in machine learning.
but also structural representation of the knowledge Definitely statistical knowledge is applied to
that a computer can understand easily. Topics machine learning through predictive analysis.
needed for understanding the methods of machine Following are the examples where Statistics can be
learning in Linear Algebra are LU decomposition, applied in machine learning.
orthogonalization, matrix operations, projections, Problem Designing
Eigen values, Eigen vectors, Vector spaces, Sympathize Data
International Research Journal on Advanced Science Hub (IRJASH) 103
no reviews yet
Please Login to review.
