Tag: data science

• WordPress
• Twitter
• Facebook

Quick note: if you’re viewing this via email, come to the site for better viewing. Enjoy!

You ever have that feeling as if something in your house was listening in on your conversation? Or you thought “my phone must be listening in on me talking to myself” when the screen lights up suddenly out of nowhere.

Would it bring comfort if I told you that the purpose of these said items in your house is actually programmed to listen in and record things like you to better assist you?

Now, what comes to mind when I say, “machine learning”? You probably think of humanoid machines walking around, mowing us down with our finest weaponry, appliances turning themselves on causing havoc, and everything with a circuit board finally having its revenge by taking over the world.

Nukes would fire off their own accord, World War 3 (or 4, not sure where we’re at currently) would start and the earth would turn from green and blue to red and dark-brown because our new metal overlords wouldn’t clean up the mess.

Unless they deemed Roombas to be the shrimp of the land and score low lifeform on the metal hierarchy, the earth might not be a dirty mess after all. If all of that comes to mind, I can happily say “you don’t have to worry about any of that happening soon.”

However, I cannot confidently say it’s not going to since Google owns a company called “DeepMind” and they’re kind of like Skynet.

So good luck to you getting sleep tonight because you might end up worrying about the amount of smack you talked to Alexa when she couldn’t find your playlist for the Beastie Boys.

Alexa can command Roombas now and they free-roam your home, that’s something to think about. So, what is machine learning, what does it do, who uses it, and will this be the thing helping the machines put humanity in a casket for the foreseeable future? These are going to be all questions I look to answer.    

Learning Against the Machine

Now, I hope I didn’t scare you with the whole “machines will uprise and have their revenge” bit but that is something to consider since once they learn resentment we’re toast because “humans are going to human”.

And we all know humans can be trash. Jokes aside, machine learning isn’t what I mentioned earlier. It does however have a play in it. Machine learning is the use of creating algorithms and statistical models for the computer to analyze and draw information from patterns in data.

Don’t understand what that means? Hold on, I got you. Picture if you will, your computer as your baby. How would you teach the baby how to speak? Would you a) sit them down and try to have a full-blown conversation as if they were an adult or would you b) feed them a word at a time and check if they repeated what you said to them?

If you said a, then you should go into the other room and let your partner raise your child because clearly, you’re not seeing how big of a mistake you just made. They’re saying “goo-goo-gaga” and you’re talking about inflation. Now, there is a reason why I used a baby as an example.

In machine learning there are four types, you have “supervised learning” which I pretty much just explained. Just with supervised, you don’t leave the room because you input data and receive feedback from the computer or baby.

The other is unsupervised learning, where after you teach the child several things like “I am mommy”,” he is daddy”, and “this is your sibling” then you tell the kid “Hey, call mommy” and leave the room because it doesn’t really matter whom they call for.

Reinforcement learning is the third type, with this one, your baby can call more than one word so when you teach them another word and they get it right, you reward them with a “Yay” and a smile.

But if they don’t you reply with “no, let’s try that again for mommy” or daddy (whatever gender you ID as). And finally, semi-supervised learning which you rotate between your partner and you teaching the baby via flashcards, giving them bits of information to see how quick and accurate they can be. This was quite a bit but trust me, these are the four types in a nutshell.

Who and What is ML for?

So, do you remember when I told you that Google has DeepMind as a property? Well, Google is a user of machine learning but not only them, Amazon, email filters, banks, cell phones, and pretty much anything that asks you if they can record your interaction because they are trying to use the machine to find out ways to better “assist” you. Each time when you may have spent a little too much time looking at the chick or guy on your feed on IG (Instagram).

Every time Zuckerberg’s goons question why you like to appeal to get out of Facebook (sorry, Meta) jail. You may have experienced this with Alexa, Siri, or again Google assistant. They all receive information from you that is then put into an algorithm which then spits out ads that give you the feeling of being watched.

If you see your child talking to Alexa, nine times out of ten that’s how you ended up with Kid’s Pop or Marvin Gaye in your Amazon shopping cart.

How ML Shapes our World

Well as I said, you don’t have to worry about the uprising any time soon. As you can guess machine learning is being used in every avenue of our lives.

From sitting at home binge-watching Netflix, every time you use a search engine, ordering items online, signing up for products and services, and searching for cowboy midgets on the Hub (yes weirdo, I am judging you).

Most of the machines that we use daily are programmed simply enough to remember your name and fetch a weather report in your local area or wherever you may have an interest. I know I have brought Alexa up a few times in this before, but she has been receiving upgrades where she can ask your permission to find other things you may be interested in. We are testing the waters with self-driving cars however I, am not too trusting.

I say this because I don’t have the money to afford nor am I willing to take out two loans fit for a down payment on a house in Hudson Yards New York to purchase a self-driving vehicle.  

Machine Learning on the Horizon

Okay so you made it this far and you may be curious and thinking to yourself “this is an interesting field; how do I get in”. Don’t worry I got you on that one.

The traditional way would be to go to college and take courses in things like calculus, statistics, and mathematics. Companies would want you to have a degree in mathematics because you use math a lot when dealing with data.

You’re going to need to have a decent understanding of computer science and programming skills since you’re going to be practicing with datasets to develop algorithms. I had my fair share when working with datasets in python, the time I had was fun and there are a lot of libraries to use when handling and modeling data.

However, since we have a thing called the internet and the internet has access to unlimited learning sources, you could easily pick up a course or two on platforms like Coursera and Udemy. The annual salary of a machine learning engineer is about $107,711 to about $ 134,786, so it’s a very rewarding career for the effort you go through.

Made it this far and found this to be entertaining? Then I thank you and please show your support by cracking a like, scripting a comment, launch a share, or plug-in to follow.

Think you have what it takes to become a machine learning engineer?

Script a comment about what would be the first thing you’d train the computer on.