According to a Forbes survey, about 13 job sectors will be fully automated by 2030, and the World Economic Forum estimates that about 75 million jobs will be displaced due to automation, but interestingly, a total of 133 million new jobs will be generated.

However, it is possible to automate some low and middle-skill jobs due to automation, but it is never possible to do data science jobs. If you want to understand why automation can’t waste data science jobs, you need to understand what data science is and how the data science ecosystem works.

Then at the beginning, you should take a little idea about data science. In simple terms, data science is a hybrid subject, and data science focuses on statistics, applied mathematics, and computer science.
We call any job lucrative when its market value, job salary, and future demand are all very good. Now you have to understand how the job market in data science?

Most of the most lucrative jobs of the present age are indeed based on data science. According to Glassdoor, a data scientist in the United States earns about $ 95k – $ 200k a year. Job positions based on Data Science are created respectively – Machine Learning Engineer, Database Administrator, Data Architect, Data Engineer, Business Analyst, Data Analyst, Data Scientist, etc.

So what is the demand in Bangladesh? Yes, it is true that the demand for data science relevant jobs is also high in Bangladesh. However, with the development of technology in Bangladesh, the demand for it is increasing and the demand will increase in the future. Many private companies in Bangladesh are hiring new data science-relevant jobs to increase their business growth rate.

You may be wondering what is the relationship between increasing the business growth rate and data science? Yes, of course. Let me clarify the matter with a small example. Suppose you search for ‘python tutorial’ on YouTube. After searching for it, YouTube may bring up many tutorial suggestions on your homepage and you may cut it after watching a video for a while.

Next time you go to YouTube, you will hopefully bring at least 1-3 Python programming videos to the homepage along with all the other videos. But what is the reason? When you search, YouTube understands that you are interested in Python. So later suggested this kind of video in front of you. Now I think the question is how these things are normal?

To discuss this in detail, we need to talk about machine learning, NLP, and data mining. ML, NLP, DM can all be called subsets of data science. I will talk about each topic in another article in a specific way InshaAllah.

I want to learn data science. But how do I get started? This question seems to be a common question of every human being. I will share some tips from my experience which I hope will be very useful for newcomers. First, you can learn by doing a master’s course in data science. If that is not possible then there are many resources available online.

Suppose you want to learn data science right now. But first, you have to learn the basics of Python or R Programming Language well. About 80{65d74b771b6ceff07eaefc19ffed56e0ab6ea89ffc0f0b38516c35f1ac414386} of people here are confused about which one to start with. The answer is whatever you like. In my opinion, the best decision is to start with Python programming. Python’s resources are available for free on YouTube.

On the Study Mart channel, you can learn basic Python programming in Bengali very well or many more English resources, including FreeCodeCamp, Krish Naik, Corey Schafer, Telusko, Edureka, Coursera.

Now let’s talk about Python, not R programming. In fact, Python’s syntax is much easier to remember, and you get a lot more Python resources from it. You have no alternative to Python when it comes to image, NLP-related work. You can easily do this through a deep learning framework based on Python, such as Tensorflow or Pytorch. Python has a machine learning library like Sklearn. Matrix libraries like Numpy, data frame libraries like Pandas are awesome in a word.

But R is not inferior in any way. And with it, almost everything can be done. However, R has a library of 2400+ statistics and 3000+ built-in functions. Which is very easy to do statistical analysis (if the target is data analysis: If you want to be a good data scientist, Python needs to be proficient in two languages. In addition to acquiring data analysis skills, you need to learn MS Excel, SQL Database, SPSS, Weka, Tableau, Power BI slowly. Tableau or PowerBi should be learned very well at least.
When your Basic Python is over, all you have to do is get a rough idea about NumPy, Pandas, Seaborn, and Matplotlib. This is because about 80{65d74b771b6ceff07eaefc19ffed56e0ab6ea89ffc0f0b38516c35f1ac414386} of the time in each data science project is spent on data preprocessing and statistical analysis.

So you need to know at least Pandas and Seaborn well. Then your job is to make an algorithm selection for you to develop your model. Since you will start with Python, you will find all the necessary algorithms in Psychic Learn. The best resource for you to learn everything can be studied mart’s Python PlayList. Everything is shown step by step in the Bengali language. Coursera, MIT OpenCourseWare, Krish Naik, Edureka, etc. as resources in English. Can follow.

After finishing machine learning you will start deep learning and N, L, P. Here too you can follow the above resources. And it is better to say that if you want to be a skilled data scientist, there is no alternative to Kaggle.

Because here you will get a lot of info from the notebook. You will also get a free dataset to practice at Kaggle and you can win thousands of dollars as a prize by participating in the competition if you want. I will tell the rest in another article. Join Study Martin our data science community Learn Data Science Smartly.

Read More:

Lessons about life before turning first big milestone you should learning

Success doesn’t require speed. It requests slowness.

Some of the most interesting facts about human behavior?

%d bloggers like this: