Let’s begin our introduction to the profession with the field in which Data Scientists work. Data Science is the science of data, which deals with the study of data, its analysis using various methods, and the subsequent transformation of data into useful knowledge. It used to be possible for humans to process data manually, but now the amount of data has become so huge that it often requires artificial intelligence to process it. Therefore, science actively interacts with machine learning, mathematics, statistics, and data analysis.

We are constantly surrounded by the results of Data Scientists’ work, for example, we watch the weather forecast every day, advertisements offer us certain products, airline services predict ticket prices, doctors can use software to predict diagnoses, and voice assistants fulfill many of our requests. All of this and many other things are managed by a data scientist. A Data Scientist is a professional who looks for patterns in large amounts of data, analyzes and stores it. The Data Scientist profession is considered one of the highest paid and most complex in the IT world.

It is worth paying attention to the fact that Data Science has become an integral part of the future. It is now actively used in startups, IT companies, and various businesses to provide the most accurate data and predictions, to be closer to the user, to automate their decisions, and to improve business margins.

The demand for Data Scientists is growing every year. For example, according to job search website Indeed, there were 29% more Data Scientists jobs for 2019.

Data Scientists are constantly looking for patterns and trends in huge data sets, using a variety of aces, techniques, and critical thinking to find practical solutions to real data-centric problems. Let’s talk more about what data scientists do.

What does a Data Scientist do?
The tasks of a data scientist are:

Finds hidden patterns and connections while examining data;
analyzes the data according to the necessary criteria that will show the effectiveness of the model being created;
visualizes the data;
programs and trains the machine learning model;
evaluates with colleagues the model in terms of economics;
identifies rich data sources, joins them with other potentially incomplete data sources, and cleans up the resulting set;
analyzes risks;
analyzes internal processes;
handles model implementation in existing infrastructures;
refines the model and monitors processes;
proposes new directions for client business development;
Develops reports and forecasts;
advises executives and the product manager based on the findings.
Thanks to the work of Data Scientist, business makes the right decisions and is ahead of its competitors, products become closer to users and people’s lives become more convenient.

In order for data scientists to excel in this field, it is often not enough for them to be effective simply in transforming masses of unstructured data into a form suitable for analysis. It is also desirable to be able to analyze the processed volumes of data themselves, to conduct factual analysis.

Data Scientist does not equal Data Engineer
You need to understand that these are not the same thing.

Data Engineers provide quality data infrastructure on projects and focus on integration, modeling, optimization, and data quality. These professionals also influence applications in an operational context in the areas of analytics, microservices architecture and operational analytics. It turns out that Data Engineers develop, test and maintain the data infrastructure, as well as deal with data: cleansing, processing and transforming it. Already cleansed data goes to analysts and Data Scientists.

The two professionals have different goals: Data Engineers work on the creation of a machine learning algorithm maintenance pipeline. And Data Scientists test hypotheses in the data system, and write algorithms. Both professionals want to make data accessible and quality and often work together. Hence the constant confusion in their duties and responsibilities.

For example, Data Scientists extract insights from data for company strategy, decision making and algorithm implementation. And Data Engineers work as a team to improve analyst productivity and be a liaison between the various participants in software development.

They say that to become a Data Science specialist you need to study all the time, but this can be said about many professions. Let’s understand what knowledge you will need in this profession if you just got into it, as well as if you are already working in a junior position and plan to grow.

Requirements for the profession
What should a novice Data Scientist know?
Programming.
A data scientist should be able to write code. A data scientist is involved in writing a model to evaluate hypotheses, analytics, or evaluate data. There is no way to do this without knowledge of the basic programming languages used in Data Science. You will need knowledge of:

Java, Hive to work with Hadoop;
Python – its basics and an understanding of how to work with it in data analysis. Also get to know Matplotlib, Numpy, Scikit, Skipy;
SQL – for data extraction;
C++ with BigARTM, Vowpel Wabbit, XGBoost;
R language, which will come in handy for calculating statistics.
Math.
The data analyst should take courses in mathematical analysis, mathematical statistics, linear algebra, and know what probability theory is. This knowledge will be useful for making predictions, working to find patterns, and building mathematical models.

In mathematical analysis, you will need derivatives, the rule of differentiation of a complex function, and gradients. Descriptive statistics, experimental planning, and machine learning will need to be studied in a statistics course. Linear algebra is needed to understand the mechanisms of machine learning, there pay attention to vectors and spaces, matrix transformations.

Machine learning.
Without it in your work is nowhere. Machine learning is needed to create new models and retrain existing models. It is also related not only to artificial intelligence, but also to genetic, evolutionary algorithms, cluster problems, and so on. With machine learning, Data Scientist’s work with large amounts of data becomes effective.

Deep Learning.
To lead machine learning projects, you will need to understand how neural networks are structured and learn the basics of deep learning.

Domain Specifics.
Understanding how a product works and creating the right model requires knowledge of the domain in which you work. Data Scientists work in all sorts of industries, the most popular of which are marketing, healthcare, and economics. If you do not have the necessary profile knowledge in advance, do not worry, you will definitely get them on the project.

English.
A must for any specialty in IT. English will help you in your work when communicating with foreign clients and colleagues in a multinational team. You will also encounter English while working with different frameworks and technologies, and in your development: a lot of technical literature is published only in English.

If you already work in Data Science, you are probably familiar with all these requirements. For experienced data analysts, of course, they are different.

Requirements for an experienced data scientist
Some experts describe a successful Data Scientist as a hacker, analyst, communicator or trusted consultant. Let’s understand what kind of skills you’ll need.

In addition to the hard skills we described above, you need to have:

Experience developing machine learning and deep learning models with Hadoop, TensorFlow, Keras, PyTorch, Scikit-Learn, Pytorch, MLLib and other frameworks;
In-depth knowledge of one of the areas of Machine Learning precedents;
Experience with SQL and BigData tools like Spark/Hive;
Experience with visualization tools Pandas, Matplotlib, Seaborne.