Machine learning in Data Science

A Presentation on
“Machine Learning in Data Science”
By
Vaibhav Kumar
Assistant Professor
Dept. of CSE
DIT University, Dehradun
Vaibhav Kumar@DIT University, Dehradun

Outline
 Data science
 Applications
 Challenges
 Machine learning
 Machine learning tools in data science
 Popular applications
 Future Scope

Data Science
 Data Science is a fast growing demand for professionals in business, public agencies and
many organizations.
 Data science is an interdisciplinary field focuses on finding the insights from data.
 These insights may be in structured or unstructured form.
 It unifies the concepts of data analysis, statistics, computer science to understand and
analyze the actual phenomena with data.
 A data scientist develops new methods and algorithms to analyze the data.

Applications
Data science may be applied in the fields like:
 Customer Analytics: To know the choices, behavior, capacity of customers.
 Fraud Analytics: To find fraudulent customers by credit card issuers and banking
industries.
 Business Analytics: To analyze the investments, revenues, profit, loss of business firms.
 Compliance Analytics: To help organizations transform their audit, risk and compliance
through data analytics.

Challenges
 Size: There is huge and ubiquitous data available everywhere and it may be available from
multiple sources. Now it is a challenge to decide what amount of data and from which
source should be taken for analysis.
 Features: So many features may be available in data. So it is a challenge to extract only the
useful feature for analysis.
 Parameters: Various parameters are available to analyze the data. So it may be a challenge
to use the appropriate set of parameters for analysis.
 Structure: Finding insights from unstructured data is a challenge.

Machine Learning
 Machine learning is a subfield of computer science which focuses to develop the
computer algorithm to learn from examples and improve the performance of a task.
 There are three broad categories of machine learning:
 Supervised Learning: Which learns from labeled examples.
 Unsupervised Learning: Which learns from unlabeled examples.
 Reinforcement Learning: Which learns from environment through feedbacks.
 It develops predictive analytics models which allow researchers, data scientists to predict
about future based on past and current data.

Machine Learning Tools in Data Science
 A list of machine learning algorithms popularly used in data science are given as:
 Regression models
 Artificial Neural Networks
 Decision trees
 Support vector machines
 Naïve Bayes
 K-Nearest Neighbour (KNN)
 K-Means
 Random forest
 Gradient boost model

Popular Applications
 Ecommerce: Companies are identifying customers based on their purchase and browsing
history to attract them on further purchase.
 Banking and Securities: These industries are monitoring customers to identify illegal
trades and future frauds.
 Media and Entertainment: Using sentiment analysis companies are providing the
products to the customers on their choices.
 Bioinformatics: Expert systems are used in identifying disease and drug discovery.
 Government: Governments are using big data analytics to study the pattern of population
for policy making
 Share Trading: Predicting the share price.
 Many other….

Future Scope
 Deep learning techniques are trend now a days in the area of data science.
 Accuracy in result and handling with large volume of data implicates the development of
new models
 Researchers are constantly developing new models in this field either by adding new
features to the existing models or by tuning the parameters of analysis.
 There will be a big demand of data scientists everywhere due to the increase rate of data
and required insights from the data.

Thank You

Machine learning in Data Science

More Related Content

What's hot

Similar to Machine learning in Data Science

Recently uploaded

Machine learning in Data Science