Today I’m going to be talking about how I learned Machine Learning in three months. If you’re in the field of computer science, Machine Learning is an extremely important skill for you to have.
But even if you’re not in the field of computer science, it is also such an important skill because it’s going to play such a big part of our future.
Machine Learning is a field of computer science, which allows computers to learn without being explicitly told to do so.
Most of you have probably experienced Machine Learning in your day to day lives already, if you guys have personal assistants at home, like Alexa or Google Home, or even when it comes to doing Google search, and getting better search results.
Besides that, Machine Learning is predicted to disrupt nearly every single sector in our world. So safe to say it’s a really important skill to learn.
How to get into Machine Learning?
First of all, there is no one size fits all approach, because everyone has a different way of learning. This is just my own personal take on it. And also a big factor into how you go about your own learning is what you hope to gain from it.
Some of you might want to be hoping to start careers in Machine Learning, hoping to get a job in Machine Learning, others might want to start a start up in Machine Learning, or just, you know, learning it for fun.
So it really depends on what your end goal is, For me personally, I learned Machine Learning when I was doing my thesis on Machine Learning.
The very first thing that you need when you’re learning Machine Learning is math. Now, I know there’s two kinds of arguments for this one, which says math is actually not really important,
because there’s so many good libraries that take care of the math like TensorFlow, and pi torch and whatnot.
And the other argument which says, you know, math is extremely important, because when you run into problems down the line, you will know how to solve it.
For me, personally, I believe that Yeah, you do need to learn math, but don’t make it don’t restrain yourself and tell tell yourself that,
Okay, I’m gonna finish learning math, and then I’ll continue on to the next step, I think you should take it on as, Okay, I’m gonna be learning this math throughout my entire journey as a Machine Learning engineer.
2. Basic Machine Learning Course
The next step is to actually take a basic Machine Learning course, which teaches you the basic Machine Learning algorithms like linear regression, and so on.
I personally did Andrew Ng course in Coursera. But there’s many other ones out there, which are just as great. The Andrew Ng in Coursera, actually was very heavily focused in math.
But that worked for me personally. But there’s many other ones out there, which are not that much focused on math, which might work for other people.
There’s actually many languages which are used in Machine Learning, but Python has kind of become the de facto language. So it is really necessary.
Now, just like with the math, you don’t have to master Python before moving on to the next step. Take it as a learning journey as well.
But learning basic Python is really helpful. And you are going to need that there are many free online resources, which are really great for you guys to learn Python that I personally learned using data camp,
4. Data preparation
This is a step which actually a lot of people overlook. But once you do your first couple of Machine Learning projects, you’ll realize that you probably spend about 60 to 80% of your analytical pipeline, just doing data preparation, data preparation is something which cannot be fully automated.
So that is the reason why it takes that takes up such a huge amount of time. Not just that, but doing good data preparation actually results in higher accuracy of your Machine Learning algorithms.
So this is why it plays such an important role. And since it’s such a big thing, I’m going to be making an entire video dedicated to data preparation in the future. So check that out when it’s out.
5. Deep Learning Library
I would recommend for beginners to start out with scikit learn because it has all the classical Machine Learning algorithms.
And when you decide to move on to maybe deep learning, you can move on to TensorFlow because that that is TensorFlow strength, because it has a lot of support for deep learning.
A lot of the times, Machine Learning engineers actually use scikit learn alongside with TensorFlow, so it’s really good to know both of them.
6. Practice, Practice, Practice
You have to reiterate what you’ve been learning in order for you to get for you to be good at it andm, One really good way to do that is by joining competitions on websites like kaggle, where you’re able to compare the effectiveness of your Machine Learning algorithm with other people.
And that’s a really, really good gauge of your own skills. One tip, which is really, really important, and which I’ve have to learn, personally, is to be patient with yourself when it comes to learning Machine Learning, because it’s not necessarily the most easiest thing to do.
But it’s something that anyone can achieve if they put their time and mind to it. And another thing is to not have any zero days. What I mean by that is not not necessarily that you have to do an entire project every single day,
but actually to spend time every day to actually learn something new that you haven’t learned before. Or just to reiterate what you’ve already learned.
At the beginning of this Article, I mentioned knowing what your own end goal is to learning Machine Learning, whether it is career or to start something for yourself, or it’s just an hobby.
Requirements For Machine Learning engineering job
So I’m going to be looking at it from starting a career in Machine Learning. So what I’m going to be doing right now is to just look at what the requirements for Machine Learning engineering job is.
So let’s take a look at this Machine Learning engineer position in Apple in Zurich. They want someone who has in depth expertise in deep learning, deep learning is actually a subfield of Machine Learning. And it makes use of neural networks.
And they also want someone who is extremely experienced in Machine Learning and reinforcement learning. Reinforcement learning, just like deep learning is actually a type of Machine Learning.
Next, they want someone who is really familiar with what is new, and what is actually happening in deep learning.
Someone who is really current with all the new kinds of technologies, you should also be able to crane and debug deep learning systems define metrics and data sets.
This is where this is where the data preparation part actually comes in. When we’re talking about defining metrics and data sets, performing error analysis, training models in a modern deep learning framework.
Modern deep learning framework usually refers to things like TensorFlow, pi torch and whatnot.
Next, they also want someone with experience with hardware specific optimization of ml models and deployment. And also excellent programming skills in Python, C and c++.
This is just to give you guys an idea of what these top tier companies actually look for in Machine Learning engineers.
For those of you who are applying for jobs in the Machine Learning field, you might realize that a lot of them asked for candidates to have either a master’s in computer science or a PhD.
And I urge you guys not to be discouraged when you see this, because the Machine Learning sector First of all, is changing daily,
because I believe that the most important thing is to have knowledge and also a lot of useful side projects related to the job that you’re applying to.
So that is what’s going to help you stand out from the rest of the candidates we are applying to the same job as you in the Machine Learning field.
Well, guys, that was how I personally got started into Machine Learning. Hope this Article was helpful.