How to become a data scientist?
13 February 2015
Data Science combines advanced maths/stats, computer science and acute problem solving skills so it is a pretty rare combination. Up until now there have not been undergraduate courses ** for data scientists with most people joining the industry having completed a Phd in Statistics, Comp Sci or Physics and then transitioning or experienced industry professionals move from Data Analytics, Business Intelligence or a similar quantitative field.
We have seen that people have transitioned very well from quantitative finance to data science as many of the technical skills are the same and quants naturally have a strong commercial side and can add immediate value within data science roles, despite the lack of ‘industry’ experience.
Similarly if you are transitioning from a top tier academic institution, have a strong background in applied stats, can code in at least one OO language and have experience with complex data analytics then you tick the key boxes and will have options available to you across both data science and quant trading world. Saying that, the ‘entry level’ market is hugely competitive within both finance and data science as companies need to invest a lot of time in all manner of training before you reach your full potential.
If you tick 2 or 3 of the above boxes and have not been lucky enough to secure a position straight from academia then there are a couple of choices – internships to gain some experience or additional training/education to improve your technical skills and open up a new career network.
In terms of training/education there are now three broad options – further study provides a strong theoretical background, structured learning and a career/social network but is expensive (Msc or Phd in Machin Learning/data Science/Stats), online study is a cheap and increasingly effective way of learning core new skills ,though you have to manage your own curriculum (Coursera, open University courses etc) or you can apply for/sign up to a data science Bootcamp/Fellowship where you will undertake an intensive and structured course and gain practical experience through experienced mentors.
The bootcamps and fellowship courses are increasingly common/popular and come in a variety of formats - from being highly selective and offering free tutelage to 12 week programs costing many thousands of dollars. Some of these offer fantastic courses, have great mentors and an awesome network of companies looking to hire. Others are profiteering from the demand for data scientists, charge exorbitant fees and then either don’t guarantee employment or tie you into working for their ‘select partners’ so they can charge a recruitment fee on top of your tuition (the latter is fair enough if your tuition is being provided for free but not if you are paying $14,000).
Different things work for different people and are different skills/experience are looked for by different firms, some need domain experience, some like raw untarnished ability.
If you would like advice on what is the best course of action for you then please do not hesitate to contact one of our experienced big data consultants – Will or Nigel.
**From Sept 2014 the University of Warwick is running a Data Science degree which is being run by a combination of the mathematical statistics and computer science faculties.