The Ultimate Learning Path to Become a

Data Scientist and Master Machine Learning

in 2019
Broadly, the learning path to become a data scientist can be divided into the following steps:

1. Getting Started: The biggest step of them all – beginning your data science journey.
This stage is all about understanding what data science is and what a data scientist role
entails. Additionally, this is where you should pick up the programming language and
tool of your choice (our recommendation is Python). This will enable you to code
through all that you learn in the coming months
2. Learning Basic Maths and Statistics: What are the core concepts a data scientist must
absolutely know? That would be statistics and mathematics. Where learning a tool will
help you perform quick calculations and generate results, you can’t truly become a data
scientist until you have a solid grasp on statistical methods (probability, descriptive and
inferential stats) and mathematical fields (linear algebra to be precise). That’s why we
emphasize on these two fields in this year’s learning path
3. Learning Machine Learning concepts and applying them: This is where things start to
hear up – you’re reading this article because you’ve been intrigued by what machine
learning can do. And once you’re done with the above points (by March start if you
follow the learning path), you will start learning the basics of machine learning. But this
isn’t just limited to theoretical concepts. We firmly believe in learning by doing, hence
we have provided some awesome projects so you can experience what a data scientist
does!
4. Some more applications of Machine Learning: Once you have a good grasp on these
basic techniques, we move along in April to more advanced topics, like ensemble
learning, random forest, boosting algorithms, and time series methods. But ML isn’t
limited to just the algorithms, you need to know nifty tricks to improve your model,
right? That’s where validation strategies and feature engineering will play a role. We also
encourage you to keep your focus on industry applications, and have hence included a
recommendation engine project in the learning path
5. Introduction to Deep Learning: Now you know these machine learning concepts, what
comes next? Deep learning of course! It’s becoming an essential part of any data
scientist’s CV these days. July should see your data scientist path lean towards
understanding neural networks and getting the hang of Keras
6. Various deep learning architectures like RNN, CNN: Follow that up with a deep dive
into advanced neural network frameworks, namely recurrent neural networks and
convolutional neural networks. These are fairly heavy concepts, hence we recommend
spending a few weeks on understanding them from scratch
7. Computer Vision Applications: Computer vision is one of the hottest fields right now
and hence we have focused a lot on this domain. We feel every data scientist should
absolutely have this on their resume since this is where a lot of the jobs in the future will
 come up. We have included a really cool project to give you a practical understanding of
how a computer vision model works
8. Natural Language Processing (NLP): No data scientist learning path is fully complete
without first going over NLP. You should focus on learning the basics at the very least,
including text preprocessing and text classification. If you’re feeling adventurous, you
can explore how deep learning works in NLP but that’s not a mandatory requirement

We have broken down all these steps month on month – so if you start following the learning
path, you know exactly what you need to follow and what you need to cover every month
starting today.

You can access the full learning path here and register yourself to start your journey today.
Our training portal enables you to track your progress after each section thus helping you
to stay on track throughout the year.

Here is an image laying out what should you do month on month to become a data scientist by
the end of 2019. If you put in all the efforts as mentioned in the learning path – you will be
well placed to get into a data scientist role before the end of the year.

We have one more gift to make your new year truly special. Join Analytics Vidhya’s CEO and
Founder Kunal Jain on January 10th for an exclusive webinar where he will elaborate on
how to get the most out of this learning path. He will discuss the roadmap to become a data
scientist in 2019. Get your questions answered and your doubts clarified by one of the
eminent personalities in this field!
Machine Learning Engineer
Machine learning engineers build, implement, and maintain machine learning systems in
technology products. They focus on machine learning system reliability, performance, and
scalability. This career path requires you to have expert-level programming skills and deep
knowledge of machine learning algorithms.
Recommended Prerequisites

 Basic Python programming: You should be able to use Python to read in data and perform basic
manipulations.

Top Skills

 Programming
 Machine Learning
 Big Data
 Cloud Computing
 System Design

Sample Projects / Problems

 Build a recommendation model to optimize customer activity and deploy it at scale in the
product.
 Iterate on existing machine learning models by engineering new features and testing
alternative learning algorithms.

Data Engineer
Data engineers design, build, and maintain data architectures for large-scale applications. They
manage the entire data lifecycle: ingestion, processing, surfacing, and storage. This career path
requires strong software engineering skills

Recommended Prerequisites

 Basic Python programming: You should be able to use Python to read in data and perform basic
manipulations.

Top Skills

 Big Data
 Apache Hadoop
 Web frameworks
 NoSQL
 Spark

Sample Projects / Problems

 Build an internal platform to automate the training of machine learning models.

 Design the data warehouse of a company to ensure high performance and easy access to
internal data.
Data Scientist
Data Scientists perform sophisticated empirical analysis to understand and make predictions
about complex systems. They draw on methods and tooling from probability and statistics,
mathematics, and computer science and primarily focus on extracting insights from data. They
communicate results through statistical models, visualizations, and data products.

Recommended Prerequisites

 Basic R or Python programming: Work proficiently with R/Python to read in data and perform
basic manipulations.

Top Skills

 Data Tools
 Regression Models
 Machine Learning
 Data Visualization
 Probability & Statistics

Sample Projects / Problems

 Design an AB test to measure the performance of a new product relative to existing options.
 Utilize causal inference techniques to disentangle the effects of various interventions on key
metrics like company revenue.

Data Analyst
Data Analysts use tools such as Excel, Tableau, SQL, R or Python to use data to answer specific
questions. Analysts must have a deep understanding of their organization’s data. This career path
requires you to be able to visualize data in ways that help guide major business decisions.

Recommended Prerequisites

 Excel proficiency: You should be able to work with data using Excel formulas, charting, and pivot
tables.

Top Skills

 Data Tools
  Excel
 Machine Learning
 Experimental Design
 Probability & Statistics

Sample Projects / Problems

 Write SQL queries to extract and clean data in order to provide key customer insights.
 Generate dashboards for key company metrics to track historical performance, correlating
movements in metrics with past interventions.

There is no such path my friend. Being a data scientist requires dedication and patience. And
with necessary knowledge and skills acquired you will be ready to pursue the journey of
becoming a data scientist.

I always advice everyone : before starting on the path to becoming a data scientist, its important
that you answer some questions below:

 Do you enjoy statistics and programming?

 Do you enjoy working in a field where you need to constantly be learning about the latest
techniques and technologies in this space?
 Are you interested in becoming a data scientist, even if it just paid an average salary?
 Are you okay with other job titles (e.g. Data Analyst, Business Analyst, etc…)?

If that's a yes, then let’s go ahead to understand your career path to become a data scientist.

I would suggest you to watch: How to make a career transition to data science

I will recommend you to do the following:

Step 1- The Math

The main topics concerning mathematics that you should familiarize yourself with if you want to
go into data science are probability, statistics, and linear algebra. As you learn more about other
topics such as statistical learning (machine learning) these core mathematical foundations will
serve as a base for your learning.

1. Probability: Probability is the measure of the likelihood that an event will occur. A lot of
data science is based on attempting to measure the likelihood of events, everything from
the odds of an advertisement getting clicked on, to the probability of failure for a part on
an assembly line.
2. Statistics: Once you have a firm grasp on probability theory you can move on to learning
about statistics, which is the general branch of mathematics that deals with analyzing and
interpreting data.
3. Linear Algebra: It covers the study of vector spacing and linear mapping between these
spaces. It is used heavily in machine learning, and if you really want to understand how
these algorithms work, you will need to build a basic understanding of Linear Algebra.

Step 2- The Key Technologies

The data science community has mainly adopted Machine learning, R and Python as its key
technologies. Let me give you a brief of it:

1. Machine Learning - Machine learning is an application of artificial intelligence (AI) that

provides systems the ability to automatically learn and improve from experience without
being explicitly programmed.
2. Python - Python is an interpreted, high-level programming language. Python allows
programmers to use different programming styles to create simple or complex programs,
get quicker results and write code almost as if speaking in a human language. Companies
are looking for this language specifically.
3. R - The R programming language is an open source scripting language for predictive
analytics and data visualization.
4. For Visualization, Tableau- Tableau is the most powerful, secure, and flexible end-to-
end analytics platform for your data. Tableau is the only business intelligence platform
that turns your data into insights that drive action. That actually very helpful if you learn
and then go for best opportunities because that is the must element companies will look
forward from your side.

Step 3- The Community

The job search for data scientist positions can take a while, its best to begin building out your
network!

One of the best ways to begin to build out your network is to attend meetups for data science!
But you don’t need to be limited strictly to data science, you should attend meetups with any
topics that are related to data science, things like Python meetups, Visualization meetups, etc.
Step 4- The Job Search

One of the most realistic ways ever to become a data scientist is to work as a data scientist.
Nothing can ever surpass the experience you gain from doing real-life projects.

Get industry exposure to enhance your skill as a data scientist. Start an internship or join a boot
camp or if you already have experience as an Analyst, then get bigger and better projects to
become an expert in Industry. And there always is the chance to convert an internship to a full-
time job!

All the best to you!

==============================-==================================

To become a Data Scientist you have follow the below steps:

Step1: Learn Mathematics and Statistics:

To be more specific you need to learn topics like Linear algebra, Calculus, Inferential Statistics
and Differential Statistics. To be very honest you got to have a great understanding about these
because if you don’t know these concepts then its useless to have good hands on knowledge
about technologies like python or machine learning. Because in Data Science we need to use all
these Mathematical and Statistical concepts in python and machine learning using libraries like
NumPy, SciPy and Pandas.

Step2: Python:

If your confident enough with Step 1 and you got to learn Python. Python one of awesome
programming language that is so easy to code that you guys will love to code in it.

Python contain many libraries that helps a data scientist to work with different forms of data and
apply different algorithms.

Step3: Learn libraries in python like NumPy, SciPy and Pandas.

For the Data Science using Python we use different packages like NumPy, SciPy and Pandas.

•NumPy:

NumPy is a fundamental package used for scientific computing in Python. It is a library in

Python that provides a multidimensional array object, various derived objects (such as masked
arrays and matrices), and an assortment of routines for quick operations on arrays, including
mathematical, logical, shape manipulation, sorting, selecting, I/O, discrete Fourier transforms,
basic linear algebra, basic statistical operations, random simulation and much more.

•SciPy:
SciPy is a scientific library in Python for mathematics, science and engineering. The SciPy
library depends on NumPy, which provides convenient and quick N-dimensional array
manipulations. The main reason for building the SciPy library is, it should work with NumPy
arrays. It comes with many user-friendly and efficient numerical practices just like routines for
numerical integration and optimization.

•Statsmodels:

Statsmodels is a Python module that provides classes and functions for the estimation of many
different statistical models, at the same time for conducting statistical tests, and statistical data
exploration. An extensive list of statistics results are available for each estimator. The results are
tested against existing statistical packages to acknowledge that they are correct.

•Pandas:

Pandas is library that provides high-performance, easy-to-use data structures and data analytical
tools for the Python programming language.

•Scikits Learn:

Scikit-learn comes with a range of supervised and unsupervised learning algorithms through a
consistent interface in Python.

The library is built upon the SciPy that must be installed before you can use scikit-learn. This
stack that includes:

NumPy, SciPy, Matplotlib, IPython, Sympy, Pandas

Extensions for SciPy care conventionally named SciKits. As such, these modules provide
learning algorithms and is named scikit-learn.

It has functions to build machine learning models like Regeration, Support vector machine,
Clustering and many more.

It also includes functions to calculate accuracy of models.

With Python you can also perform data visualization using some pictular libraries like
MatPlotLib and SeaBourn

MatPlotLib:

It is a Python Library that supports 2D and 3D graphics.

It is used produce publications like Histogram, Power Spectra, Bar Chart, Box Plots, Pie Chart
and Scatter Plots with just few lines of code.
It easily integrates with Pandas Data-Frames to make visualization quickly and conveniently.

SEABOURN:

It is built on top of MatPlotLib and introduces additional plot types.

It also makes MatPlotLib visualization Elegant.

It is mostly used create Complicated Plots with ease.

Heatmap uses visualization which can create with Seabourn using just one lik=ne of code.

The IDE that has changed the python programmers can leverage code with documentation and
live output all in the same document known as NoteBook.

NoteBook:

Here Data Scientist can present their report in a story telling kind of format as multiple blocks of
code can be run with output of each block of code displayed right below it.

It works as magical organizer for data scientist.

You may be curios to see the state of python libraries in the area of future of Data Science.

Pythons development for Deep Learning libraries for Googles TensorFlow and frameworks like
Theano have enabled the Data scientist to built artificial neural networks.

To make our point on why python for data science you can view the Kaggle survey for the same.

Step 4: Learn Machine Learning:

Here we use machine learning library in python in order to train the machine to make decisions
with great approximations in order to avoid failures.

Step 5: Keep Practicing the above 4 to be a master.

For further reference.

Thank you.