Data Science in Madrid
How much time a new Data Scientist have to wait to find a Data Science Job?
To answer this question, I analyzed more than one thousand LinkedIn profiles of Data Scientist who live in Madrid. The analysis focus on the academic backgrounds of the profiles and how much time past from the end of their Data Scientist Studies and his first Job as Data Scientist. Following this analyzes I made a model to know the probability of find a job in a period like the people who has the same academic background.
The Data
Profiles: public information of LinkedIn of more than 1000 people self-defined as Data Scientist living in Madrid.
Academic Data of the Data Scientist in Madrid
Only half of the Data Scientist has studied something related to Data:
The Data Scientist has always another academic background different than Data Science:
We can group the no-data-related background in Technical (Physics, Engineering), Mathematics (Math, Statistical),Humanities (Social Science, Philosophy, Psychology), Economics (Economy, Administration, Financial, Accounting), Marketing and Natural Sciences. And the distribution is like this:
The types of the degrees are:
Something to be noticed is that Data Scientist posted in average 4 items in his education section:
Measurement of Job Search Time
This project goal is predicting the time a person who studied something related with Data Science have to wait to find his first job in Data Science. For this this I calculated the time past for the end of the first Data related Studio and the start of the first job in Data Science, based in the LinkedIn information.
As I said, at first only the half of the profiles has a Data-related Study and now I find that only a quarter of the sample has studied Data and then find a Job in Data Science. In other words, a quarter of the profiles has a job of Data Science before of studied something related with data. This is because they are people who studied technical fields and know a lot of Data Science.
Relation between academic background and the time of searching for a Data Science Job
It’s not the same looking for a job of Data Science when you have a technical background than when you have a humanities background. It´s not the same have and Engineering degree than a bachelor’s degree. It´s not the same have a lot of academic Background, then only have studied Data Science. But this data impacts in the time of searching for a job? I believe it does. And that’s why I do this research.
In the following graphics we can see the possible relation between the Background and the time of looking for a job.
It´s curious that the people who has studied Marketing and Economics has and average more time waiting for a job in Data Science. Probably because they can use Data Science in their own fields without the title.
This graphic shows the relation between the number of Data Science studies and the time of finding you first job. The people who have more Studies waits more to find his first job in Data Science.
The Model
The range of days was between -800 days and -15, expressed in negative numbers because it’s a subtraction. I decided binned in three groups: 0.8, 1.56 and 2.33 years to find a job. And the data is distributed like this:
After testing several models, the best was a Random Forest Classifier. It was not a perfect model, but it had the best accuracy and F1 result. You can see a Confution Matrix:
Project link: https://github.com/dfmoscoso23/linkedin_project