Data Science Techniques, Tools and Predictions.

For conference Proceedings Published IEEE

Shivakumara B Prof. Hetal Rana

Dept. of Artifictial Intelligence and Data Science Dept. of Artifictial Intelligence and Data Science
MVJ College of Engineering MVJ College of Engineering
Bangaluru, India Bangaluru, India

ABSTRACT :- Almighty created human being with increasingly relying on data science techniques and tools
to gain a competitive edge, optimize operations, and
numerous wants and needs which makes them
drive innovation. This introduction sets the stage for
associated with their own data, choices and
exploring the evolving landscape of data science
preferences[1]. Data science incorporates various
techniques, tools, and predictions, highlighting their
disciplines such as statistics, mathematics, computer
significance and impact on decision-making processes . A
science, and domain knowledge to analyze complex data
data Scientist is a specialized person in data science,
sets[2]. Its applications span across industries such as -
which will not only analyze the data but also makes use
finance, healthcare, marketing, and more, where data-
of machine learning algorithms for the future prediction
driven decision-making is crucial for success. Earlier than
of events. It is a Combination of three different fields
data science, we had statisticians. These statisticians
that is mathematics, statistics, and computer science.
skilled in qualitative evaluation of records and
organizations hired them to research their standard
overall performance and income. With the arrival of a
computing technique, cloud storage, and analytical
equipment, the field of Computer science merged with
information. This gave birth to statistics science.
Predictions in data science encompass forecasting future
trends, identifying anomalies, making classifications, and
optimizing processes based on historical data
patterns[4]. As data science continues to evolve, fueled
by technological advancements and the proliferation of
data sources, its impact across industries is poised to
deepen, shaping the way organizations understand and
leverage data for strategic decision-making and Fig:1 Data Science Overview
innovation.
OBJECTIVE OF DATA SCIENCE :- The
INTRODUCTION :- Data Science is the primary objective of data science is to extract meaningful
systematic workflow of extraction, preparation, analysis, insights, patterns, and predictions from large and complex
visualization, and maintenance of information. It is an datasets [2]. By employing a combination of statistical
interdisciplinary field which is based on scientific analysis, machine learning algorithms, and data
visualization techniques, data scientists aim to uncover
methods and processes to gain knowledge from raw
valuable knowledge that can inform decision-making,
data[1]. Popular tools in the field include programming
optimize processes, and drive innovation across various
languages like Python and R, as well as libraries and domains. Ultimately, the goal of data science is to
frameworks such as Tensor Flow, PyTorch , scikit-learn, leverage data-driven insights to solve real-world
and pandas. With the exponential growth of digital problems, enhance efficiency, and create value for
information, organizations across various industries are organizations and society as a whole .Another major
concern is to correct the drawbacks depicted in the the relationships between different variable of
previous projects or mishandling of data. The principal different groups.
objective of Data Science is to find interesting patterns
within data[1]. So, for finding patterns, a Data Scientist i. Time Series Analysis: Measurement is done based
must scrutinize the data thoroughly by using various on time series for the variables of data sets.
statistical techniques like data extraction, wrangling and
pre-processing to analyze and draw insights from the Data Analysis Tools
data. After doing that they will make predictions from the
data. The main Objective of a Data Scientist is to make The major role of Data Scientists Is to make decisions
meaningful conclusions from the data [3]. By using these which are done by analyzing and handling lots of
conclusions, companies are able to make smarter unstructured data and structured data. As stated in
business decisions. paper there are numerous big data technologies that
DATA ANALYSIS METHODS :- Exercise have been advanced and classified into data
and follow good process in collecting the data by using processing concepts.[4] So to handle such a large
various qualitative and quantitative approaches. Data amount of data programming languages and tools are
Analysis [8] can be divided into.[2] needed for data scientists to analyze that data and do
their work appropriately. In this article we will explore
a. Textual analys : is which can also referred as some available tools for data science which is useful
data mining it is to arrange the data into large data for analyzing data and generate predictions.[7]
sets using mining tools. The main aim of textual
analysis is to map the data into business data using As stated in [16][8] the growing need in the market
business intelligence tools. for Information technology professionals demands
from data analytics. It becomes considerably essential
b. Descriptive Analysis: It is to interpret, model to deploy the various data analytics tools in
and process the previous collected data which can accordance with rising need of
be done instatistical analysis. society. Below is the list of top 10 of data analytics
tools which are open source and as well as paid
c. Inferential Analysis: In which we can investigate
versions to improve the performance and learning of
various inferences from the same data various
samples. the Ssystem.

d. Diagnostic analysis: These methods are to

investigate the statistical analysis and find the cause
for why it happens.

e. Predictive analysis: In this analysis we try to

predict what can happen by using statistical data. For
example in day to day life how the person does save
on his predictable earning income.

f. Prescriptive Analysis: This form analysis is used Fig: 2 data Analysis tools
to collaborate all the previous analysis reports to Excel: This is product of Microsoft suite and developed
decide what decision could be taken based on current under Microsoft Office family for performing
situation. mathematical, statistical and analytical operations. Excel
is the essential and important entity as analytical tools
g. Factor Analysis: This analysis speaks about how the used in various organizations. It plays an important role
variables form the relationships within the data set. by analyzing the complete user requirements and précis
in way which is useful to users. It also used for business
h. Discriminant Analysis: This analysis is used to find analytics which helps in presenting of automatic
relationship detraction. It can be used in creating budget TOOLS FOR DATA SCIENCE
sheets for personal and business purposes as come up in The major role of Data Scientists Is to make decisions
[8].
which are done by analyzing and handling lots of
R Programming Language: It is free unstructured data and structured data. As stated in
paper [5] there are numerous big data technologies
software programming language and reinforced R
that have been advanced and classified into data
foundations for statistical computing. The R Language
processing concepts .So to handle such a large
is widely used data analysists by mining the data and
amount of data programming languages and tools are
statistical information. R is used as analytical tool
needed for data scientists to analyze that data and do
which can be used in various ways to extract and
their work appropriately. In this article we will explore
present the data of the many organizations as stated
some available tools for data science which is useful
in [8].
for analyzing data and generate predictions. Table 1
summarizes availabletools of Data Science.[1]
Python: Python is developed by Guido van Rossum
created it in the early 1980s, dynamic all-purpose
purpose high programming language supports both Data Science Contributions for the future
structured and object oriented programming. It Data Science comprehends many advances
stated in [8][16] Python also rich in library & open technologies like Artificial Intelligence, Internet of
source and considered for functional & structured Things (IoT’s), Deep learning, machine learning and so
techniques which is used to implement various on. With the advancements in technology demands to
tasks. Python can assemble in & from any platform incorporate & implement the statistical, mathematical
such as Mango DB, JSON, SQL, server and many more. and logical reasoning concepts. Proper mechanisms
need to stratagem to make the organizations to
handle the data with operative use. There are
numerous reasons to give for which we need data
SAS: With reference to [8] it is abbreviated as
science operations to be performed in business
Statistical Analysis System developed in between the like.[10]
year 1980’s & 1990’s by SAS institute. SAS is a
programming environment for managing the data and • Organizations how they mishandle the data
analytical operations. This programming language is • Data protection to formulate the regulations &
used to manage the data from various sources can be policies, a surprising incline in data growth
analyzed which can be serve to client profiling and • Much demand for the data scientists
future opportunities. This SAS modules used for Web, • Natural Language Processing (NLP) will be used
Social and market analytics. for information retrieval
• Data purgative should be computerized
Apache Spark: As come up in [8][16] Apache • Much need to improve business intelligence
spark was created in university of California in the • Used to predict sports, whether, banking
year 2009 AMP lab of barkely., Spark rummage-sale sector, stocks and shares
for micro- batching for real time streaming by an- • Need much improvement in social media
applications.
alyzing large amount of data from various resources.
Like Hadoop it also works with the system by Conclusion
distributing the data over various clusters and
Know a day’s data science becomes as a
processesthem in parallel.
mandatory field which coordinates between
multi disciplines like mathematics, statistical
approaches, mathematical methods, logical 56.12 (2013): 64-73.
reasoning, intelligence algorithms and machine  [5] Bejjam, Suvarnamukhi &
learning practical’s. All these fields correlate to Seshashayee, M.. (2018). Big Data
access the data from various business or Concepts and Techniques in Data
organizations and make use of them in effective Processing. International Journal of
means. These effective use of data leads to Computer Sciences and Engineering.
perform proper decision making to grow business  Ethem Alpaydin (2004). Introduction to
Machine Learning, MIT Press, ISBN 978-0-
further on the basis of customer chooses and
262-01243-0.
satisfaction. Hence we can conclude that rise of
 Stuart Russell & Peter
data science field can demand more positions of
Norvig, (2009). Artificial
data scientists to grow in each organization. At Intelligence – A Modern
last we focus on how successful carriers can be Approach. Pearson, ISBN
built in the field of data science. The main beauty 9789332543515.
of this field it used to grow all businesses.  https://data-flair.training/blogs/data-
science-tools
REFERENCES  Blei, D.M., & Smyth, P. (2017,
August 15). Science and Data
 Russell, Stuart J., and Peter Norvig. Science. Proc Natl Acad Sci U S
Artificial intelligence: a modern A, 114(33), 8689-8692.
approach. Malaysia; Pearson http://doi.org/1073/pnas.170207
Education Limited,, 2016. 6114
Gruson, D., Helleputte, T., & Rousseau P. (2016,
 Nicolae, Bogdan, et al. "Park, Yoonho.
July). Data science, Artificial Intelligence, and
Leveraging Adaptive I/O to Optimize Machine Learning: Opportunities for Laboratory
Collective Data Shuffling Patterns for Big Medicine and the Value of Positive Regulation.
Data Analytics. IEEE TRANSACTIONS ON Clin Biochem, 69, 1-
PARALLEL AND DISTRIBUTED SYSTEMS. 7.http://doi.org/10.1016/j.clinbiochem.2019.04.01
PP (99) pp: 1-13." (2020). 3.
 Islam, Mohaiminul. "Data Analysis:
Types, Process, Methods, Techniques
and Tools." International Journal on
Data Science and Technology 6.1
(2020): 10.
 Van Der Aalst, Wi l. "Data science in
action." Process mining. Springer, Berlin,
Heidelberg, 2016. 3-23.
 Rani, Bindu , and Shri Kant. "An
Approach Toward Integration of Big
Data into Decision Making Process."
New Paradigm in Decision Science and
Management. Springer, Singapore,
2020. 207-215.
 Dhar, Vasant. "Data science and
prediction." Communications of the ACM
[Type text] Page 5