Artificial Intelligence and Machine Learning.

P

Proverbial

Banned
Joined
Mar 7, 2018
Messages
1
Reaction score
0
It looks like all the R&D in this field happens in the USA, China, Russia, South Korea and so on. Are there any companies in SA that also does research is quite up to date with the technology?
 
I've heard of a few companies that use it in SA - mostly market research trying to get insight from client data. Very little on improving the state of the art or even doing large scale ML. The universities don't appear to be big on it either - the focus there is often on very abstract aspects of AI as opposed to ML.
 
cguy said:
I've heard of a few companies that use it in SA - mostly market research trying to get insight from client data. Very little on improving the state of the art or even doing large scale ML. The universities don't appear to be big on it either - the focus there is often on very abstract aspects of AI as opposed to ML.
Click to expand...

That was my feeling on the universities as well. I was in the process of enrolling for a Masters in CS and I have an interest in ML & AI specifically. 90% of the topics I was presented with were incredibly theoretical and abstract. Nothing practical. It really killed my interest.
 
The field is growing in SA, I've noticed the financial sector paying a specific interest. (and oddly also in the security sector).
 
I'm studying some data science courses online, and they will both cover machine learning later on in the syllabus.

Coursera has a 9 course data science specialization offered by the Johns Hopkins university. Seems good so far, the second course is a lot more practical which is good. Machine learning is coming up later and I have no doubt it will be very practical.

I'm also doing a Microsoft Professional Program in Data Science. That also has a machine learning component, although I think it will be a lot more practical and less theoretical than the Coursera one.

I think it is important to have knowledge of both the theory and practice. After doing these data science courses, I want to a mathematics for machine learning specialization (also on coursera). I think without the theory, you are just a technologist.
 
Ancalagon said:
I'm studying some data science courses online, and they will both cover machine learning later on in the syllabus.

Coursera has a 9 course data science specialization offered by the Johns Hopkins university. Seems good so far, the second course is a lot more practical which is good. Machine learning is coming up later and I have no doubt it will be very practical.

I'm also doing a Microsoft Professional Program in Data Science. That also has a machine learning component, although I think it will be a lot more practical and less theoretical than the Coursera one.

I think it is important to have knowledge of both the theory and practice. After doing these data science courses, I want to a mathematics for machine learning specialization (also on coursera). I think without the theory, you are just a technologist.
Click to expand...

Depends what you want to achieve. cguy can confirm, but I suspect it's rather unlikely that you'll get very far on coursera courses, when interview at boutique investment firms that run algos and similar. So at some point, you need a more formal qualification.
 
I plan to do an internal move rather, and start working in the field where I am, rather than moving to a new place.

Plus once I've studied these things, I should be able to get acceptance to study an MSC in either data science of operations research. Considering I already have a BSc (Hons) in computer science.
 
Ancalagon said:
I plan to do an internal move rather, and start working in the field where I am, rather than moving to a new place.

Plus once I've studied these things, I should be able to get acceptance to study an MSC in either data science of operations research. Considering I already have a BSc (Hons) in computer science.
Click to expand...

You don't need to have much/any experience in ML or AI to get access to a MSc in the field - honours is sufficient. It'll likely help some, but again, the topics you'll be presented with are incredibly abstract vs even what the Coursera courses teach.
 
I've looked at some of these online courses focusing on machine learning and/or AI and wasn't that impressed. At this stage I'm focusing on statistics and will at a later stage decide on a direction.
 
SlinkyMike said:
Nonsense.

Come work in my business, I hate PhDs and we do plenty of ML, Data science, etc.

DISCLAIMER: only scrappy punks who regularly overflow with passion allowed.
Click to expand...

Super. What's your business, what do you do and where are you? Dead serious here.
 
Messugga said:
You don't need to have much/any experience in ML or AI to get access to a MSc in the field - honours is sufficient. It'll likely help some, but again, the topics you'll be presented with are incredibly abstract vs even what the Coursera courses teach.
Click to expand...

I did artificial intelligence as an honours level course, but that was a long time ago. Actually what worries me about some of these data science masters programs is that Coursera seems to cover a lot of the same stuff. I think I'd only really benefit from a MSc that does not involve coursework - in other words, 100% thesis.

Too early for me to decide though. I need to finish these courses first and then later on in the year, I can transition to working in the field. Then I can decide what further studies I might need.

xrapidx said:
I've looked at some of these online courses focusing on machine learning and/or AI and wasn't that impressed. At this stage I'm focusing on statistics and will at a later stage decide on a direction.
Click to expand...

They are light on statistics. I think they are useful as addons to other education that you may have. Such as, if you already have decent statistical knowledge, they can give you more practical experience in how data science is used.

Plus they are very convenient. You watch the lectures when you want to watch them, no driving to a university required.

A single Coursera course is not worth much, but I think a specialization adds up. In fact, I did some Masters level engineering courses at Wits - I'm pretty sure that some of them were less in depth than a lot of coursera courses.
 
xrapidx said:
I've been wondering how far I'd get with a HP Microserver I have lying around. :D
Click to expand...

uh dude... /look of disapproval

Nefertiti said:
Do you have any experience with Kaggle (https://www.kaggle.com/)? I learn best with examples so I have been looking for real life ones and came across this.
Click to expand...

Kaggle is great.

One thing though, and this pertains to your use case is that the data are all prepared and formatted perfectly - one of the hardest parts about any analysis is getting the damned data in and ready to be used. Many tricks of the trade exist in this area and you need to be able to do this to do anything meaningful, so Kaggle might leave you open to a bit of a shock when you try to do real world stuff.
 
SlinkyMike said:
Kaggle is great.

One thing though, and this pertains to your use case is that the data are all prepared and formatted perfectly - one of the hardest parts about any analysis is getting the damned data in and ready to be used. Many tricks of the trade exist in this area and you need to be able to do this to do anything meaningful, so Kaggle might leave you open to a bit of a shock when you try to do real world stuff.
Click to expand...

Im used to working with getting unformatted data into a mssql databases so i know pain :o I have no ML or AI experience however so kaggle looks like a good way to get that type of experience without having to worry about the data too much.
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X