I finished the Coursera Data Science Professional cert, which is an IBM course that uses the Jupyter Notebooks and Python in IBM Watson. I used Watson enough to run out of free time and had to sign up for the single user account to finish up. Well... is this such a good idea?
My account billing estimate is now at about $132 after about 2-3 weeks. That doesn't panic me at the moment as they have this $200 credit for new upgrades. I have had a problem with the relationship between my MACOS and the Jupyter Notebook, which the IBM people don't seem very interested in answering. Someone on Stack Overflow directed me to a Jupyter site in SO so I logged the problem there, then noticed that a couple hundred other problem reports had been languishing there for up to 2 years. Hmm. This is a bad sign, I can tell.
IBM Watson/Jupyter notebooks seem to me to only provide value when groups are collaborating and the IBM Cloud has value when an Enterprise needs a host for their application. Otherwise, they don't seem to add value for an individual who is just trying to learn the trade and push out a few studies.
Am I missing something or is that about it?
Also, can anyone provide a human scale interpretation of the IBM Watson billing structure? Billowing clouds of vCPU hours and instantiations charges. I think that, given the charges that I've accumulated so far, and the $200 credit for the upgrade, it all means $200/month.
I can't answer your question about costs but my one experience with Watson was in grad school and was quite disappointing. It would be last on my list of stats tools to consider for just about anything unless I can be convinced our application was not suited for the platform. IBM knew what we were doing and donated the time so I doubt that will happen.
Thanks a lot for sharing your experience with us. Always good to talk ;)
So let me explain a couple of things. First, the jupyter notebook part of IBM Watson Studio is only a tiny little fraction of the whole functionality. And the only reason why I’m using it in the course is to give learners a runtime where everything needed is pre-installed and pre-tested.
Watson Studio is a very large integrated toolbox for applied data science including automated data pre-processing, machine learning, deep learning in all sorts of variants. For the non-programmers there are visual editors for each and every task. It allows for automated hyper-parameter tuning, scaling and deployment. I’m on purpose not explaining all this in the courses since I’m emphasizing on open source and don’t want to do marketing. Maybe I should have explained this a bit more in detail.
I’m stating in front of every programming assignment that the parts we are using from IBM Watson Studio are fully open source and you can use your own environment as well. But still you can benefit here if you want to use more CPU cores or main memory. Here you can see how much of improvement it can give you in the free tier (just skip forward to 19:03)
In the free tier (no credit card is needed) you can scale up to 21 CPU cores and 84 GB of main memory.
So it is not only about collaboration, but having a set of similar tools in place is really helpful to get AI into production, I’ve tried to summarize this here:
The only thing what we can’t do is supporting you in fixing your own environment. We’ve tried that and believe me, it is not feasible with a set of courses which has 120K+ of learners.
That’s why some course mentors might have pointed you to use stackoverflow.com to fix your environment.
I can't follow how you've spent $132 in 3 weeks since this means nearly 600 hours of continuous runtime. Maybe you didn’t stop the kernels in between? I’ve tried to make this clear in the following video, so you have unlimited time for running an Anaconda-notebook and 50h per month for free on Apache Spark (currently working on it to give you even more free credit):
And since you don’t need to swipe your credit card ever, you are perfectly save.
Btw, let me introduce myself. I'm Romeo and the course lead of the courses of the "IBM Advance Data Science Specialization" on coursera (nopanic.ai/datascience) including the courses
"Fundamentals of Scalable Data Science" => http://coursera.org/learn/ds/
"Advanced Machine Learning and Signal Processing" => http://coursera.org/learn/advanced-machine-learning-signal-processing/
and "Applied AI with DeepLearning" => http://coursera.org/learn/ai/
You are talking about the "Coursera Data Science Professional cert", are those the courses you are referring to?
There is a set of less advanced courses from IBM giving you the "Coursera Data Science Professional Certification", please let me know which courses you are referring to.
Hello, Romeo. I haven't ignored your response. I've visited the URLs you cited and I've been doing some additional homework.
First, I was an aircraft engine control systems engineer, then engineering program manager and Black Belt before retiring. I also took the GE internal curriculum for being certified as an "Analytic Engineer". The curriculum was principally Coursera courses, like courses offered by Johns Hopkins and Andrew Ng's course on machine learning and neural networks. I recently put myself through the 9 course 'Data Science Professional Certification' on Coursera, which is authored by IBM. Yours, I think that you are saying? My purpose was to dust off as I'm far to young to be mothballed. I was surprised to learn how much Python has advanced and how big the market of services has become in only a few years.
I clearly have much to learn. I got the 2 week freebie from Tableau as it seems to be a handy backpocket tool. The courses you cite above seem like they should be on my immediate future to-do list also.
I'd like one more word of advice, if you don't mind. I'm taking courses and, knowing that I'm starting over in a new field, searching for positions more like entry-level. I'm also doing studies to publish both as blog posts and to send as unsolicited research to agencies. Does this sound like a good launch strategy?
Thanks for your reply above.