Building a datascience bootcamp for Africa

[emsibbz]

Over the last couple of months I have been enrolled in an international data-science bootcamp. Having had a keen interest in the field for quite some time, while I have found the opportunity great, I have noticed a lot of flaws in how some bootcamps seem to operate.

For starters, 6 months is nowhere near enough to train a non technical person to a data scientist, unless that person is extremely committed, learns quickly and probably has some prior stats/programming experience and knowledge. Secondly, the cost of bootcamps is very high. Particularly, international bootcamps. You can easily expect to pay over r100k, there is limited mental support. Its one thing to have a great curriculum, its another to keep students dedicated and motivated especially when they encounter the inevitable imposter syndrome and mountains of self doubt...and some work can be done to improve the curriculum and how the mentorship model works (in a lot of bootcamps you have a mentor helping to guide you through your capstone projects and the course material)

That being said, I really wanted to start getting more input from data scientists around Africa to build a better understanding of what they believe are the most important skills to acquire in order to successfully transition to a data science career. I want to use this research to build a new local alternative. In order to aide in completing this research I created a few questions in a form and wanted to find out if anyone is interested in answering a few of the questions I have?

 

[cguy]

So after 2 months doing a data science bootcamp, you decide that you would rather run your own bootcamp than work as a data scientist?

My suggestion would be to finish, and then try and work in the field. This way you can get a better sense of how your education matches up to the particular job/field you end up working in, where it works and where it does not for said job, but also and where you ideally would have liked to work, but could not for some reason you have yet to discover.

Asking what are important skills for a data scientist is like asking what the best programming language is (it’s C BTW ). The skill set for those who call themselves data scientists today seem to run from excel wranglers and sklearn duct-tapers to PhD level maths/stats/cs/etc scientists and everything in between.

My inclination is to focus on the fundamentals (maths/stats/programming/ML), which runs contrary to the idea of bootcamps (well, those that don’t require STEM education at least), but to my point above, this doesn’t mean that people without a formalized education can’t be effective in many situations (for the same reason that the person integrating a payment system doesn’t need a PhD in Computer Science and a person with a PhD in computer science may not be suited for such work).

 

[konfab]

Figure out a way to cram a 3 year degree in stats and computer science into 6 months and you have a data science boot camp.

 

[SlinkyMike]

Somewhat related: https://zindi.africa/competitions

...African Kaggle analogue I guess you could say. Some interesting problems and decent prize money!

(Sorry for mild derailment but I think this is a good place to post this.)

 

[emsibbz]

So after 2 months doing a data science bootcamp, you decide that you would rather run your own bootcamp than work as a data scientist?

My suggestion would be to finish, and then try and work in the field. This way you can get a better sense of how your education matches up to the particular job/field you end up working in, where it works and where it does not for said job, but also and where you ideally would have liked to work, but could not for some reason you have yet to discover.

Asking what are important skills for a data scientist is like asking what the best programming language is (it’s C BTW ). The skill set for those who call themselves data scientists today seem to run from excel wranglers and sklearn duct-tapers to PhD level maths/stats/cs/etc scientists and everything in between.

My inclination is to focus on the fundamentals (maths/stats/programming/ML), which runs contrary to the idea of bootcamps (well, those that don’t require STEM education at least), but to my point above, this doesn’t mean that people without a formalized education can’t be effective in many situations (for the same reason that the person integrating a payment system doesn’t need a PhD in Computer Science and a person with a PhD in computer science may not be suited for such work).

Thanks. It has been quite a lot longer than 2 months. I have been enrolled in my current bootcamp for a little over 8 months, prior to that I had enrolled for some introductory courses in another bootcamp and had been self learning for quite a while. I think the interest stems from the fact that I generally have been an entrepreneurial person. I love learning data science, but I also have an entrepreneurial side that I cannot ignore.

I agree with the fundamentals running contrary to the idea of bootcamps, the focus on fundamentals is at times superficial but the idea is that they get you to a position where you can get a job and they assume that you are self-driven enough to fill in the gaps on the job.

Would you be keen to answer a few more questions- can I send through a few questions?

 

[emsibbz]

Figure out a way to cram a 3 year degree in stats and computer science into 6 months and you have a data science boot camp.

I think the point shouldn't be to replicate what a student would learn in University, some argue that bootcamps are stronger on the practical side of things. They get you to a level where you can get employed and then you learn the rest on the job (presumably you have some level of understanding of the fundamentals either prior to joining the bootcamp or after- and you continue to build these up on the job)

 

[cguy]

Thanks. It has been quite a lot longer than 2 months. I have been enrolled in my current bootcamp for a little over 8 months, prior to that I had enrolled for some introductory courses in another bootcamp and had been self learning for quite a while. I think the interest stems from the fact that I generally have been an entrepreneurial person. I love learning data science, but I also have an entrepreneurial side that I cannot ignore.

I agree with the fundamentals running contrary to the idea of bootcamps, the focus on fundamentals is at times superficial but the idea is that they get you to a position where you can get a job and they assume that you are self-driven enough to fill in the gaps on the job.

Would you be keen to answer a few more questions- can I send through a few questions?

Sure, send away, although note that I haven’t been based in SA for awhile.

 

[emsibbz]

Somewhat related: https://zindi.africa/competitions

...African Kaggle analogue I guess you could say. Some interesting problems and decent prize money!

(Sorry for mild derailment but I think this is a good place to post this.)

Thanks, will definitely be doing that

 

[emsibbz]

Sure, send away, although note that I haven’t been based in SA for awhile.

Sure, send away, although note that I haven’t been based in SA for awhile.

Thank you, I appreciate it -> https://techmentor.typeform.com/to/SSJKSc

 

[Mosgi]

Thanks. It has been quite a lot longer than 2 months. I have been enrolled in my current bootcamp for a little over 8 months, prior to that I had enrolled for some introductory courses in another bootcamp and had been self learning for quite a while. I think the interest stems from the fact that I generally have been an entrepreneurial person. I love learning data science, but I also have an entrepreneurial side that I cannot ignore.

I agree with the fundamentals running contrary to the idea of bootcamps, the focus on fundamentals is at times superficial but the idea is that they get you to a position where you can get a job and they assume that you are self-driven enough to fill in the gaps on the job.

Would you be keen to answer a few more questions- can I send through a few questions?

I guess the question would be the purpose, there is nothing wrong with creating something that can get you to a point where you are competent enough to work within a greater team where you aren't necessarily the lead on the problems you are solving.

Would it therefore not be worthwhile to consider what would be entry level for a consulting firm who gets people with basic BSC or other qualification and getting them through a focussed application course ? This allowing them to therefore not start at zero.