Im underground, back where it all started.
Sitting at the hidden cafe where I first met Mike.
Id beenstudying in my bedroomfor the past 9-months and decided to step out of the cave.
14-months into my machine learning engineer role, I decided to leave and try it on my own.
It’s free, every week, in your inbox.
Add it up and you get about 3-years.
So as it stands, I feel like Ive done a machine learning undergraduate degree.
Thats the thing with knowledge.
Learning is non-linear (not a straight line).
You may study for an entire month and feel like youve made zero progress.
Then seemingly out of nowhere, a discovery appears.
If you want an example of how we fool ourselves, did you catch the error?
It seems I still forget how to spell.
But enough about me.
Yours might be similar or you might be starting out today.
If youre getting started, this article is for you.
If youre a veteran, you’ve got the option to offer your advice or critique my ideas.
Lets get into it, shall we?
Ive even created my own.
Theyre all remixes of the same thing.
Instead of worrying about which course is better than another, find a teacher who excites you.
Learning anything is 10% material and 90% being excited to learn.
How many of your school teachers do you remember?
My guess is, regardless of what they taught, you remember the teacher themselves more than the material.
Dabble in a few resources, youre smart enough to find the best ones.
See which ones spark your interest enough to keep going and stick with those.
It isnt an unpleasant task to learn a skill if the teacher gets you interested in it.
Every new shiny framework which comes out, every new state of the art model, Im onto it.
Often Ill catch myself trying to invent a problem to use whatever new tool is on the market.
A classic cart before the horse scenario.
A chefs entire work centers around two tools, the controlled use of fire and a knife.
This is embodied in the best programming advice Ive ever received: learn the language, not the framework.
Too many models live and die within Jupyter Notebooks.
And deployment, as in getting your models into the hands of others, is hard.
But thats exactly why I shouldve spent more time there.
Dont be afraid to make something simple.
A basic front-end which someone can interact with is far more interesting than a notebook in a GitHub repo.
No really, how?
Youll want your models to work faster, better.
To achieve this, youll need to research alternative ways of doing things.
Youll find yourself reading research papers, replicating them and improving upon them.
Im often asked, how much math should I know before I start machine learning?
To which I usually reply, how much walking should I know before I go for a run?
As a side note, Ive just ordered theMathematics for Machine Learning book.
Im going to be spending the next month or two reading it cover to cover.
Having read the free text online its more than enough to cover the fundamentals.
Skill before certificates
Ive got online course certificates coming out of my ass.
I got caught thinking more certificates equals more skills.
I optimized for completing courses instead of creating skills.
Because watching someone else explain it was easier than learning how to do it myself.
Everything I learned for an exam, Ive forgotten.
Everything I learned through experimenting, I remember.
Now, this isnt to say online certifications and courses arent worth your time.
Courses help to build foundational skills.
But working on your own projects helps to build specific knowledge (knowledge which cant be taught).
Before youaddsomething, ask yourself, have I sucked the juice out of what Ive already covered?
Thats bullsh*t. Youve got the internet.
you’ve got the option to learn anything.
The internet has given rise to a new kind of hunter-gatherer.
The following path isnt set either.
Its designed to be a compass rather than a map.
Its all accessible online.
Lets lay some foundations.
An excerpt of the 2020 Machine Learning Roadmap.
Note: This curriculum is heavily focused on code-first, Python code in particular.
It also neglects mobile or embedded gadget development.
However, it contains more than enough resources to get an outstanding grounding in the field.
My main goal would be to build more things people could interact with.
Alongside these, Id go through:
Theres a lot here.
And of course, these would be shared on GitHub.
Again, after going through these, Id consolidate my knowledge by building a project people can interact with.
An example would be a web system powered by a machine learning model.
Good news is, it’s possible for you to get both of these yourself.
I created my ownAI Masters Degreeas a form of accountability and structure.
you’re able to do something similar.
Its similar to mine but includes more software engineering practices.
Im still not quite sure how to do this.
I say self-driven here because all knowledge is largely self-taught.
Rather than hand-feed knowledge, the role of an instructor is instead more to excite, guide and challenge.
Friends, does exist a platform which allows students to:
Create their own curriculums (e.g.
Has this been tried before?
Building and creating is exhaling.
Dont hold your breath.
Balance your consumption of materials with creations of your own.
Your shared work is your new resume.
GitHub and your own blog.
Use the other platforms when needed.For machine learning projects, a runnable Colab notebook is your minimum requirement.
Whats missing?
Though, since were talking about code and math, it either works or it doesnt.
Knowing this, the contents of the materials you choose doesnt matter as much as how you learn it.
Thisarticlewas written by Daniel Bourke and was originally published on hiswebsite.
you’re able to read ithere.
