This text has been translated from Google Translate.
Apparently, to be a DataScientist, you have to be good at math. And that's probably true!
Does it take a brain designed in a particular way to train neural networks?
A DataScientist designs learning models. In a company, IT is used a lot to plan resources of all kinds. And there is a lot of data that is stored on the subject. What if we could learn from all this data to predict what tomorrow will bring?
This is the topic! How many people are going to visit the store tomorrow? How many packages will this warehouse ship? Instead of looking at yesterday's figures, Data Scientists build tomorrow's indicators directly.
The model can look like many things! We talk about machine learning when the program is able to improve itself in production (QCM Machine Learning ).
After training his model, the Data Scientist may just have a linear regression model or a polynomial. If he made a workbook, he will surely have a decision tree.
There are 3 main families of algorithms:
Neural networks have been around for a long time. But before, we didn't necessarily have enough computer power to use them. Deep Learning is everything we said above, but with neural networks. I recommend the MCQ on the basics of Deep Learning.
If you want to listen to a DataScientist, I recommend you listen to Guillaume from Voodoo. You will be in his universe really very quickly!
A DataScientist uses quite a few tools. His workspace will surely be in R.
He will probably code in Python (QCM) because he uses SciKit ([QCM scikit-learn](https://welovedevs. com/app/fr/tests/python-scikit-learn).If your company is in the Google cloud (GCP), it will surely use Tensorflow!We also have a [Tensorflow QCM](https://welovedevs.com/ app/fr/tests/tensorflow).
Before the arrival of artificial intelligence, there was already BI, Business Intelligence, business intelligence in French. And we had integration developers working on ETLs (Extract Transform Load), which are data pipelines and data warehouses. Those who transformed this data into information or indicators and dashboards were called BI consultants or decision-making consultants.
These trades still exist! The first became Data Engineer, the second Data Analyst. The first may be a Java backend developer who has become a fan of Hadoop, Spark etc... The second may have gone to business or management school!