Data science - data science, how to become a data scientist from scratch date sighting

The monthly non-fiction magazine Harvard Business Review, published by the Harvard Business School, called Data Scientist "the most desirable profession of the 21st century." Interest in this area in Russia is growing every year. People are attracted by the variety of tasks solved by employees of this profile, and high salaries. More often, managers, web analysts, and programmers become data scientists. But it happens that doctors of sciences are also involved in data processing.

What is data science

Literally translated, data science is the science of data. It allows you to process large amounts of information (Big Data), visualize the results of research, use the findings in further work.

There are two stages of the process:

• data. At the first stage, data is collected, stored, and processed with the selection of useful information from the general array. It takes up to 80% of the working time of specialists.
• Science. Using the methods of statistics, optimization, machine learning, information is analyzed, formulating useful patterns for subsequent use.

By solving real problems in practice, the data scientist must find a solution that benefits a particular project.

Data science, as applied research, includes:

• putting forward a hypothesis;
• planning of experimental activities;
• evaluation of the result obtained and its applicability to the solution of the problem.

Arrays of information for analysis are heterogeneous. There are 3 types of data:

• structured;
• semi-structured;
• unstructured.

working with data

The last type is the most difficult, since Excel tables are not suitable for its digital processing, but special programs are required.

How to learn data science

Such a profession cannot be obtained in universities and colleges. To be successful in the field of data science, you need a set of practices and skills.

There are 2 training options:

• viewing lectures on the Internet;
• online courses.

After completing the course, the student receives a certificate of their successful completion. Lectures do not provide an opportunity to validate learning new skills.

Special literature will help to master Data science from scratch. For example, Cathy O'Neill's book "Data science. Insider information for beginners. Including the R language." In it, the author introduces the reader into the interesting world of analytics in a fascinating language, talking about algorithms, financial modeling, and data visualization.

Which experts work with data

At the stage of processing unstructured information and converting it into databases, ELT specialists work. These include:

• Data Engineer, whose task is to ensure the integrity and safe storage of information databases;
• backend developer - responsible for maintaining databases in a healthy form;
• the database architect plans to store the collected information.

When analyzing arrays of information, it is required to extract the maximum of useful data. These goals are implemented:

• data analyst (data analyst) - processes information to solve a problem using statistical methods, experiments, gives forecasts for the future;
• data scientist - receives information from various sources to establish patterns and develop business;
• Bl-analyst - using ready-made solutions, is engaged in their visualization;
• Ml-specialist - knowing programming languages ​​and putting forward hypotheses, develops analysis algorithms.

Specialization Data Scientist

The full functionality of a Data scientist depends on the direction of the enterprise where the specialist works.

Main job responsibilities:

• collecting information from different channels for further analysis;
• forecasting models of the customer base, its segmentation to promote specific products;
• study of sales effectiveness;
• analysis of various risks;
• preparation of periodic and one-time reports with visualization of the results obtained and forecasting of indicators for the future;
• detection of fraudulent schemes for dubious transactions.

A good specialist in this industry differs from a beginner in the ability to identify logical chains in the general array of information, offering management the best business solutions.

Essential Skills a Data Scientist Can't Do Without

The profession exists at the intersection of mathematics, physics and computer science. In addition, knowledge of statistics and basic programming is required.

Basics of programming

Requirements for applicants:

1. Knowledge of a programming language for statistical information processing (R or Python), a structured query language (SQL) for working with databases.
2. Knowledge of statistical tools (Tableau, MATLAB), application of statistical methods in practice.
3. Knowledge of linear algebra, mathematical analysis, probability theory. They will come in handy if developers decide to create their own implementations or complement existing ones.
4. Understanding of machine learning techniques, most of which are implemented in Python or R.
5. The ability to process information in a fragmented form (with gaps, string or date formatting) and convert them into databases.
6. Visualization of results and communication of information. This skill is especially relevant when bringing information to a wide audience. It is important to understand the principles of information coding and data translation rules.
7. Knowledge of the area of ​​activity of the company. For example, in medicine it is necessary to deal with the types of diseases, forms of treatment, names of drugs.

Important qualities

When compiling a resume for the position of a data scientist, the applicant focuses on personal traits that may be useful in future work. Among them indicate:

• analytic mind;
• perseverance;
• purposefulness;
• perseverance;
• concentration;
• conscientiousness;
• the desire to bring what has been started to the end;
• sociability;
• the ability to convey the meaning of complex concepts and principles in simple words.

To increase the chances of finding a job in the chosen company, during the interview, try to convince the employer that the qualities indicated in the questionnaire are really inherent in you.

Pros and cons of the profession

Every job has positive and negative sides. A person who decides to study in the field of Data science must weigh the pros and cons so as not to regret his choice in the future.

Advantages:

• the demand for data processing specialists is constantly growing, so the number of various courses for training professionals in this area is increasing;
• high level of wages;
• the possibility of constant self-development, the use of advanced technologies in the field of programming.

With all the charms of the profession, there is a fly in the ointment.

Flaws:

• the Data scientist profession requires an analytical mindset, not everyone can learn its basics;
• the application of well-known methods does not always work the first time, the search for a solution can take a lot of time and cost a lot of painstaking work.

The persistence in achieving the goal and the competence of a data scientist will help to overcome difficulties and achieve research results.

How much does a data scientist cost

In June 2019, the New.HR agency published data from a survey of Data scientists. The "net" salary of workers in this profession in Moscow ranged from 113 to 305 thousand rubles a month, depending on the length of service.

Factors affecting the salary level of data scientists:

• work experience in the specialty;
• depth of ongoing research;
• geographical location of the company - in the capital the salary is higher;
• Proficiency in English makes it possible to get a job in a foreign company, where salaries are higher.

A person who is just starting to work as a data scientist can earn good money, regardless of the region of residence.

Actions that add value to the data scientist

A data scientist can add value to his services by following a few simple rules.

salary increase

5 steps to a pay rise:

1. Follow the news in the field of data analytics, be interested in trending areas, fill in the missing knowledge.
2. Take part in seminars, round tables, conferences on professional topics. Don't be a passive listener. Voice new ideas. Gain recognition among peers.
3. Upgrade your skills in your area of ​​expertise. Experts in a particular field are more valued by employers.
4. Create a team of like-minded people. Own startup experience is welcome when applying for a job.
5. Learn to speak a language that business understands.

For passive analysts, there is an easier way - to constantly monitor the labor market by submitting a resume to a company with a higher salary level.

Can AI put analysts out of work?

Initially, computers were perceived by many as a pile of iron. But over time, machines learned to think, manage processes, and freed a person from routine work. The history of artificial intelligence dates back to 1950, but even today, in most industries, a computer cannot completely replace a person. Data science is one such area.

Analysts need to learn new technologies and apply themin their work. Artificial intelligence will help them process massive amounts of information, but will not offer alternative solutions that take into account the influence of various factors.

Where to study to become a Data Scientist - a specialist in big data

It is better to start learning data science from scratch right after graduation. Few universities train data scientists. Professional analysts are trained according to special programs by a number of educational institutions. Among them:

• Higher School of Economics (HSE) - Faculty of Computer Science - master's program in Russian and English;
• Moscow Institute of Physics and Technology (MIPT) - Faculty of Innovation and High Technologies - Master's degree;
• Lomonosov Moscow State University (MGU) - Faculty of Computational Mathematics and Cybernetics - Master's program for 2 years;
• St. Petersburg State University (SPbGU) - 2-year master's program in English "Business Analytics and Big Data".

Children can study full-time: on a paid and free basis. The master's program involves obtaining a second higher education.

There are non-profit continuing education courses for people of all ages. You can study on them after passing the entrance exams, having overcome the required threshold for points. The term of study is 2 years.

List of courses for training specialists in the field of Data science:

• Yandex School of Data Analysis;
• Technopark Mail.ru and Bauman Moscow State Technical University (emphasis on training system engineers);
• Center for Computer Science (Yandex with Jet Brains);
• Petersburg Data School (E-Contenta).

There are many commercial courses on data analysis on the Internet. Their cost is 100-200 thousand rubles. The term of study is from 2 to 8 months. Transfer money for studies, making sure that the chosen courses are not a scam that breeds “dummies”.

You can learn data analysis remotely at the Institute of Internet Professions Netology. Depending on the Data Science section, the cost of courses ranges from 25 to 200 thousand rubles. Full information is available on the official website https://netology.ru/ .

The Open Data Science company trains beginners and creates joint analytical projects. It organizes free international conferences on topical issues and areas of development, holds competitions among data scientists.

To get an online education, a person needs a laptop with Internet access and a desire to learn.

Video tutorials, books, online lectures on this topic are available on the network.

Training program

The curriculum is approved by the course developer. It defines the list of disciplines and the time allotted for their study.

Future data scientists study:

• the basics of programming in Python;
• linear algebra;
• mathematical analysis;
• fundamentals of statistics and probability theory;
• machine learning;
• neural networks;
• data engineering;
• management;
• business fundamentals.

Subjects depend on the future specialization of the student. The training program is not aimed at the theoretical study of textbooks, but at acquiring practical skills in data analysis, entering the profession.

Student Requirements

The conditions for admission to higher educational institutions are determined by the local acts of the university. There are no age restrictions. It is necessary to choose a university, submit documents on time, pass entrance examinations (testing, exams, interviews).

Prospective students should have a basic knowledge of mathematics and the basics of programming.

What languages ​​are worth learning

To work in the field of scientific data processing, you should learn programming languages. Python and R are common among beginners. Analysts also use Java, SQL, Scala.

Python

The language was created in 1991, the name python is common in Russian. Has a free license.

Advantages:

• ease of study;
• reliability;
• wide distribution guarantees developer support.

Among the shortcomings, users note the appearance of error messages due to the dynamic typing of the language. For the narrow purposes of statistical analysis, it is inferior to the R language.

R

The R programming language appeared in 1995. The license is free.

Pros:

• a variety of specialized open source packages;
• availability of a large number of statistical functions;
• vivid data visualization.

The R language is not suitable for general purpose problems due to statistical specialization.

It is inherent in the slowness of information processing.

Place of work

A data scientist is in demand wherever data is used to solve specific problems. This can be a financial institution that uses scoring systems when lending to individuals, or a transport company that schedules buses based on an analysis of passenger traffic.

work after training

• Major Internet companies. Having settled in a similar firm as an intern, you can gain experience in the field of data processing for career growth. Employees are provided with official employment, full social. package bonuses.
• Analytical divisions of enterprises of various industries. This group includes banks, audit firms, telecommunications operators, retail networks. Sberbank, one of the first financial giants in the country, used the services of data scientists. Working in the research department of a large company, an experienced specialist can make a personal contribution to its development by suggesting ways to solve old problems based on data analysis.
• Startups in the field of data science. Consulting firms are recruiting a team of data scientists. For high-quality consultation of the client, a comprehensive approach to the analysis of his activities, the formulation of proposals for business promotion is important.

The business community needs competent data scientists. Therefore, there are always many vacancies for analysts in the field of finance, telecommunications, marketing and other areas on the labor market.

How to Work in Data Science Without a Degree

According to statistics, only 1% of professional analysts are PhDs. It is not necessary to defend a doctoral dissertation in order to identify patterns in the analysis of an array of information.

It is more important for specialists to have practical experience in data processing and be able to present the results obtained to management in an accessible way.

Career and prospects

Getting the position of a data scientist is prestigious in itself, as it requires a thorough theoretical background and work experience in several professions. The analyst's opinion is considered by the company's management when making key decisions, which increases the weight of the position in the eyes of colleagues.

In the coming years, interest in the profession will only increase, which, given the shortage of specialists in this industry, will lead to an increase in salaries and an increase in the prestige of employees of analytical departments.

Interesting facts about the profession 

The profession of a data scientist allows you to look beyond the horizon without losing touch with reality. For the period 2015-2018 the need for such specialists in Russia has increased by 7 times.

5 facts about data scientists:

• Four out of five practicing data scientists are men.
• Women between the ages of 18 and 24 make up 40% of female professionals employed in this field.
• More than 60% of vacancies and applicants are located in Moscow.
• 90% of applicants have higher education.
• Only 5% of vacancies contain a freelance offer.