You are not a Data Scientist
The IT industry is undergoing rebranding: engineers are turning into architects, and deep learning can compete in popularity with cats. The time has come for data geeks to become data scientists.
Who are the “data scientists” and what they are really doing, Grigory Bakunov will tell at our career meeting “How Data Scientist Lives” .
')
I bet that you have colleagues or acquaintances who proudly call themselves Data Scientists? Do not take to heart, but most of them have nothing to do with Data Science. To call yourself a scientist is to really do science, practice scientific methods. You put forward a hypothesis, support it with the results of experiments and after its proof / refutation go ahead or carry out new iterations.
The science of data is applied science. And the task of any applied scientist is to create models, methods, algorithms that are of practical value.
These things are very important, as they can predict future results from relatively small input data. In some cases, your models are nothing but black boxes: you cannot explain where the forecast came from, but you have already proven the accuracy of this data.
Thus, in order to preserve the purity of the concept of “Data Science”, here are a few statements that will help you understand that you are not a Data Scientist:
- You have a rich expertise in business intelligence. You spent a lot of time predicting the past by doing time series analysis of historical data. This is not the science of data - you rarely conduct experiments, your ability to predict is deceptive.
- Programming experience in Hadoop, R, Python, Octave, Mathematica and Matlib - Data Scientist tools. The ability to use tools does not give you scientific influence.
- A degree in mathematics, statistics, econometrics does not give you the right to call yourself a data scientist. We hope that you have learned how to use descriptive and prognostic methods while maintaining an understanding of the basic theory. But data science is an applied discipline that focuses on a particular data subject area, so most likely, you will not get enough real-life experience by pursuing a bachelor's degree.
- Promoting the role of large, medium, small, and any data like future intellectual entrepreneurship looks appropriate in your resume, can be a way to start a conversation with a geek or entertain friends at a party. You do not become a scientist from this.
“The eight-week course at Cousera or visiting a science camp for Data Scientists makes you the same scientist as yesterday’s golf lesson makes me a professional athlete. The theory “live and learn” and the thirst for constant self-improvement is simply self-deception.
- The subject matter expert and master of Excel, you create incredible charts, graphs and crazy tables. Again, these irreplaceable skills do not make you a scientist.
- You recently acquired a Data Science platform from SAS, IBM or Microsoft and without due experience after reading the instructions, watching 10 introductory videos and passing a five-day course, you believed that you were ready to create predictive / explanatory models of subject data by simply moving the widgets of algorithms to the canvas and pressing a button "Learn". You are not a Data Scientist at all, but simply a dangerous subject.
Who are the “data scientists” and what they are really doing, Grigory Bakunov will tell at our career meeting “How Data Scientist Lives” .
')
I bet that you have colleagues or acquaintances who proudly call themselves Data Scientists? Do not take to heart, but most of them have nothing to do with Data Science. To call yourself a scientist is to really do science, practice scientific methods. You put forward a hypothesis, support it with the results of experiments and after its proof / refutation go ahead or carry out new iterations.
The science of data is applied science. And the task of any applied scientist is to create models, methods, algorithms that are of practical value.
These things are very important, as they can predict future results from relatively small input data. In some cases, your models are nothing but black boxes: you cannot explain where the forecast came from, but you have already proven the accuracy of this data.
Thus, in order to preserve the purity of the concept of “Data Science”, here are a few statements that will help you understand that you are not a Data Scientist:
- You have a rich expertise in business intelligence. You spent a lot of time predicting the past by doing time series analysis of historical data. This is not the science of data - you rarely conduct experiments, your ability to predict is deceptive.
- Programming experience in Hadoop, R, Python, Octave, Mathematica and Matlib - Data Scientist tools. The ability to use tools does not give you scientific influence.
- A degree in mathematics, statistics, econometrics does not give you the right to call yourself a data scientist. We hope that you have learned how to use descriptive and prognostic methods while maintaining an understanding of the basic theory. But data science is an applied discipline that focuses on a particular data subject area, so most likely, you will not get enough real-life experience by pursuing a bachelor's degree.
- Promoting the role of large, medium, small, and any data like future intellectual entrepreneurship looks appropriate in your resume, can be a way to start a conversation with a geek or entertain friends at a party. You do not become a scientist from this.
“The eight-week course at Cousera or visiting a science camp for Data Scientists makes you the same scientist as yesterday’s golf lesson makes me a professional athlete. The theory “live and learn” and the thirst for constant self-improvement is simply self-deception.
- The subject matter expert and master of Excel, you create incredible charts, graphs and crazy tables. Again, these irreplaceable skills do not make you a scientist.
- You recently acquired a Data Science platform from SAS, IBM or Microsoft and without due experience after reading the instructions, watching 10 introductory videos and passing a five-day course, you believed that you were ready to create predictive / explanatory models of subject data by simply moving the widgets of algorithms to the canvas and pressing a button "Learn". You are not a Data Scientist at all, but simply a dangerous subject.
Source: https://habr.com/ru/post/293308/
All Articles