For #transformationtuesdays I’m sharing one of the most common use cases for Machine Learning that a lot of people who are young or young at heart can relate to – computer games.

For today’s example, it’s how AI Bots can be better than humans at playing Starcraft.

Why Starcraft? Because its complexity and the multitude of factors to consider that affect gameplay is closer to the number of factors a human brain considers in real life when making decisions, as compared to other games like Chess for example, where players take turns.

Why is this important? Because implementing Machine Learning Algorithms on games like Starcraft will allow us to learn more about how we can build better Bots that will be applicable to the real world.

Read on for more details and to see the actual use case.

Before I show you the use case, allow me to give you an idea just how complex it is to build one of these Bots that can beat human players in Starcraft.

To give you some perspective, Google’s computer processors that crunch through Machine Learning Algorithms like Google Translate are called TPUs (Tensor Processing Units).

To beat world-class chess players, the AlphaZero algorithm needs 4 TPUs.

On the other hand, AlphaStar, the algorithm designed to beat professional Starcraft players, uses a whopping 32 TPUs! Thats a lot of processing power.

I’m sure by now you’re interested to find out more about the use case, so I won’t delay it any further. Check it out now.

Why Starcraft is the perfect battleground for testing AI

How it worked is they trained the Machine Learning algorithm by having it watch 44 days worth of human game replays for each of the 3 Starcraft races – Protoss, Zergs, and Terrans, for it to identify patterns and develop its own gameplay strategies.

It got so good, that what they’re looking at now is how to reduce the required processing power to accomplish the same tasks.

Taking on challenges like this teaches us humans on how we can be better at creating new AI tools that work in a multitude of environments and with a lot of variables, which are closer to the number of complexities and factors humans actually experience and deal with in the real world.

Ending Note

I hope you found this to be of value.

If you have any questions, reply with a comment. I’d love to hear from you.

And if you think this is helpful and you’d want to get more, then subscribe for updates, and get a free copy of my book – The Business Optimization Blueprint.

The post How AI Bots can beat Humans at Starcraft using Machine Learning (Use Case) appeared first on Robbie Agustin.

It’s #futurefridays and today we’ll be talking about the basics of Artificial Intelligence.

Previously, I talked about Natural Language Processing, and before that, Computer Vision.

But today we’ll tackle another field of Artificial Intelligence called Machine Learning.

Wikipedia defines Machine Learning as:

Machine learning (ML) is the scientific study of algorithms and statistical models that computer systems use to perform a specific task without using explicit instructions, relying on patterns and inference instead. It is seen as a subset of artificial intelligence. Machine learning algorithms build a mathematical model based on sample data, known as “training data”, in order to make predictions or decisions without being explicitly programmed to perform the task. Machine learning algorithms are used in a wide variety of applications, such as email filtering and computer vision, where it is difficult or infeasible to develop a conventional algorithm for effectively performing the task.

Let me show you some examples of how Machine Learning is being used in real-world scenarios.

But first, I would recommend you watch this video from Accenture that gives a really simple explanation of what Machine Learning is, in 1 just minute.

AI 101: What is Machine Learning?

Video © Accenture

Now that you have a basic understanding of what Machine Learning is, allow me to give some examples of it in action today.

Waze

Have you ever wondered how Waze is able to tell if traffic is heavy heading towards your destination, and it is able to forecast arrival times and recommended routes?

Well, that’s Machine Learning in action. It learns the patterns based on how much time users are spending on each road on the map and at specific times of the day, week, month, etc.

Chat Bots and Virtual Assistants

I’ve talked about this in another post talking about Natural Language Processing as this is actually a combination of Machine Language and NLP.

Check out the detailed article here: Chat Bots Making Phone Calls on Your Behalf using NLP AI (Use Case).

Email Spam Filtering

Did you know that your email software uses Machine Learning to determine which messages to filter which messages are Spam compared to those which are not?

And when you find something in the Spam folder that isn’t supposed to be there, when you flag it as “Not Spam” you’re essentially teaching the Machine Learning algorithm what to do next time.

Again this technology is a combination of Machine Learning and Natural Language Processing working in tandem with each other.

Face Recognition like in Social Media

Have you ever wondered how Facebook is able to auto-detect who the people are in a photo that you upload even if you haven’t tagged them yet?

Here’s an example of how Facebook uses this technology.

This is actually Machine Learning working in tandem with Computer Vision.

This can also be used to identify emotions in faces, like Anger, Sadness, Happiness, and so forth.

Object and Character Recognition using Computer Vision

This is pretty similar to face recognition but can also be utilized for other objects, like telling apart humans from animals, or telling apart cats from dogs, or squares from circles.

This can also be used to identify characters like numbers and letters. And when combined with Natural Language Processing, can be used to identify context and meaning.

Here’s an example of a Lego Brick Sorter using Computer Vision and Machine Learning.

Forecasting and Prediction Models like Product Recommendations

Have you ever wondered how social media sites are able to post ads that are relevant to stuff that you have been researching about?

That’s machine learning in action, where the algorithm is able to recognize your buying intent based on previous purchases and recent searches.

Machine Learning can also be used to identify patterns that can be used for forecasting, such as when NASCAR race cars should pit for fuel and tires, and when crops will grow in certain areas, given the weather and other factors.

Here’s an example of using AI to predict Heart Attacks before it happens.

Ending Note

I hope you found this to be of value, and gave you a sense of what Machine Learning is and where it is being utilized in real-world scenarios.

If you have any questions, reply with a comment. I’d love to hear from you.

And if you think this is helpful and you’d want to get updates on the next article, subscribe for updates, and get a free copy of my book – The Business Optimization Blueprint.

The post Artificial Intelligence Basics: What is Machine Learning? appeared first on Robbie Agustin.

For #wisdomwendesdays I’ll be helping shed some light on the fears that have been brought about by the coming of Artificial Intelligence.

Though I’ve mentioned in one of my previous articles that:

“Though we all know that technology is continuously improving. But we shouldn’t put all our trust in it just yet. Yes, Google’s AI has nearly twice the IQ of Siri, but a 6-year-old child is still smarter than both. And even if it does improve further, 75% if customers, like you, still say that they would want to interact with a real person more as technology improves. child.”

I’d still like to talk about this a bit more, especially on why AI is still considered “Primitive.”

There’s another article from the IEM Community that talks about this in detail, which I think is worth a read, because I completely agree and share the same sentiments.

As for my view, I believe that Artificial Intelligence tools are like the technological equivalent of Savants – they’re good at one thing, but suck at everything else.

Let me give you a couple of specific examples.

The Lego Sorting AI Machine that I’ve shared previously has been trained to sort Lego Bricks, but if you give it other types of tasks such as sorting different types of recyclable materials, or tools in a workshop, it won’t really be able to understand what to do.

Another example is the AI used to predict Heart Attacks a year before it happens. While we all know that health is wealth, you can’t expect the same tool to predict stock market trends to make you financially wealthy.

So in a nutshell, Artificial Intelligence tools are like the Savants of the tech world. They are better than humans at certain tasks, but they are far from becoming super intelligent beings capable of replacing humanity.

If anything, Artificial Intelligence will be tools that humans use to enhance our own abilities, to allow us to be more productive, and help us achieve greater things.

Ending note

I hope you learned something new today.

If you want to learn how to do these things, and you think a career in Artificial Intelligence is for you, even if you’re just a beginner, I would recommend you check out the IBM Applied AI Professional course, which you can check further for details.

The post Why AI (Artificial Intelligence) is still considered “Primitive” appeared first on Robbie Agustin.

For today’s #transformationtuesdays we’ll talk about an innovation in AI (Artificial Intelligence), particularly Natural Language Processing.

You know when there are big corporate bosses or entrepreneurs who get to delegate tasks to their Assistants such as booking dinner reservations, hair stylist appointments, and other tasks that would require the Assistant to contact other people to complete the task?

What if you had the power to do all that, without having the need to hire a living breathing Assistant, and all you have to do is command your mobile phone to do those kinds if tasks for you. Wouldn’t that be awesome?

Well that’s exactly what I’ll be showing you today.

You see, Google has been busy developing the next generation Google Assistant, where they have injected it with Google Duplex technology to give it the power to make Phone Calls on your behalf.

Some of you may already know this, but for those who aren’t aware of the existence this technology yet, I’ll share with you (1) The Demo when it was launched, so you know how it works, and (2) Where it is being applied now, and where it will be used next.

And for those who want to know what goes on inside these kinds of Chat Bots, and where Natural Language Processing and AI comes into play, then I’ll break down the bits and pieces for you so you’ll get a better understanding.

Let’s go!

If you haven’t seen the Google Assistant + Google Duplex Launch and Demo yet, now is the time to see it. Otherwise, you’re missing half of your life.

Isn’t it amazing that it’s as if you have this little creature living inside your mobile phone whom you can command to make dinner reservations on your behalf?

That launch video was in May of 2018. Fast forward to today, as per this article by Android Authority in October 2019 (which also talks about how you can activate it so give it a read), Google Duplex is now available in 48 States in the US, and has recently been rolled-out in New Zealand.

Now if you’ve read thus far, you’re probably curious about how Chat Bots like these work, and where Natural Language Processing comes into play.

It’s actually a combination of the following:

1. Voice Recognition – How this works is that when you speak into your phone’s microphone, it converts your spoken voice into waveforms, which are basically images of sound. That’s how software algorithms are better able to tell the difference between words like “where,” “were,” and “we’re.” And then it turns that information into Text.
2. Natural Language Processing – I’ve written a separate article that defines NLP, along with various examples, that could demystify this for you. But in essence, it takes the text information from the voice recognition piece, identifies the context, then provides an appropriate response, in text.
3. Text to Speech – This is pretty much like Voice Recognition but in reverse. So it takes the text information from the Natural Language Processing piece and converts it to a waveform that when played, sounds like a voice conversation.

So in its basic form, a Chat Bot can start out as a very elaborate script, where you take the most common things that a human would say, and the Chat Bot would process that information and provide a pre-programmed response.

But if you add a Machine Learning algorithm to it, you can give it more information in the form of training data, it will learn how to adapt to new conversations it hasn’t encountered before, and adapt accordingly based on what it has experienced in the past i.e. based on the training data.

Training data can be composed of information that’s already been labeled by humans.

For example, if the input is a question like “Do you think I should go to the gym today, or sleep in?” The training data can have information on sentiment (e.g. anger, disgust, fear, happy, sad, surprise, neutral) and what the common human responses are to a question like this.

And then you add Voice Recognition to the input side, and Text to Speech on the output side, make all of them work simultaneously, and you’ve got yourself a Chat Bot that has the ability to make Phone Calls on your behalf.

Ending Note

I hope you found this to be of value, and gave you a sense of how Natural Language Processing is being applied in real-world scenarios.

If you have any questions, reply with a comment. I’d love to hear from you.

And if you think this is helpful and you’d want to get updates on the next article, subscribe for updates, and get a free copy of my book – The Business Optimization Blueprint.

The post Chat Bots Making Phone Calls on Your Behalf using NLP AI (Use Case) appeared first on Robbie Agustin.

For #wisdomwendesdays I’ve been asked by someone trying to learn Data Science what the difference is between Variance and Standard Deviation in terms of purpose.

For the sake of the uninitiated, I’ll share an article that shows the simplest explanation of what Variance and Standard Deviation are.

Now back to the purpose of each, the easiest analogy I can give you is like how the Red and Blue Pills work in the movie – The Matrix.

Let me explain.

Standard Deviation = Blue Pill

Standard Deviation is easier to visualize in the real world because it is in the same unit of measure as the Mean or Average, the Median, the Minimum and Maximum values, and so forth.

So if your purpose is on how you’re going to visualize your data and plot it on a graph for you to make decisions based on what you see, then it would be better to use Standard Deviation.

Variance = Red Pill

Now if you want to do some higher-level analysis, only then will you appreciate the benefits of utilizing Variance, which is essentially Standard Deviation, squared.

An example I can give you is when you would want to combine algorithms or mathematical models to form a new one, you can easily combine Means, but not with Standard Deviation.

What you can do is square the Standard Deviations to get the Variances, and then combine the models.

After which if you would want to get the Standard Deviation of the resulting model in order for you to create a visual representation via a graph or chart, simply take the Square Root of the Variance.

Ending note

So what do you think? Do you agree? Can you think of any other situations where Standard Deviation or Variance is better to use than the other?

The post What is the difference in purpose of Variance and Standard Deviation? appeared first on Robbie Agustin.

For #transformationtuesdays I’m sharing how AI is Transforming the way you can determine if you will encounter any health issues in the near future – specifically Heart Attacks.

Read on to see how they made it work.

What they did was to take a lot of ECG scans, where the researchers already knew the outcome for each i.e. which ones ended up getting a heart attack and by when, and which ones did not get a heart attack.

The researchers then fed the data into a Machine Learning algorithm for it to train on that data.

The results are astounding, as the AI was then able to predict upcoming heart attacks and other heart-related diseases better than real-life Cardiologists e.g. those who said the ECG scans were clear, but the patients still ended up getting heart attacks.

Source: https://www.rt.com/news/473282-ai-predicts-heart-attacks/

Ending note

I hope you learned something new today.

If you want to learn how to do this, and you think a career in Artificial Intelligence is for you, even if you’re just a beginner, I would recommend you check out the IBM Applied AI Professional course, which you can check further for details.

The post Use AI to predict Heart Attacks a year before it happens (Use Case) appeared first on Robbie Agustin.

For today’s #futurefridays I’m going to answer a question that confuses a lot of people trying to learn Data Science and Machine Learning.

The question is “What’s the difference between Classification and Regression?”

Let me give a shot at this with a simple explanation and example.

Think about the output that you want to achieve.

If you have a dataset, and the output you want to get are labels or categories, then it’s classification.

However if the output you want to get is a numerical value based on computations done on the dataset, then it’s regression.

Let me give you a specific example.

Facebook uses both classification and regression in their algorithms.

The difference is their where they are used.

Classification

When you post on Facebook and upload an image, and it has faces of people who are Facebook users, the algorithm is able to determine who these people are, and tag their names accordingly. What’s happening there is called Classification.

Regression

Now when people like, comment, and share the post, the algorithm uses that to determine how “viral” the post is, and if it should be shown to other people’s feeds or not. This is usually based on count data and a corresponding weight for behavior or action. What’s happening there is called Regression.

Ending note

If course this is an oversimplification of Classification and Regression. There’s more to it than that such as contextual analytics, quality of content, determining if it’s a human or a bot, etc.

But my objective here was to demystify things and help shed some light on the topic.

I hope you learned something new today.

If you want to learn how to become a Data Scientist, some of the best learning tracks I’ve seen are the Data Science learning paths from DataCamp.

You can start learning today even for free. Check it out! Click the link below.

Data Science at DataCamp

The post DEMYSTIFIED: What’s the difference between Classification and Regression? appeared first on Robbie Agustin.

For today’s #futurefridays, the car enthusiast in me has its juices flowing.

With all they hype about the new Cybertruck, people either love it or hate it. And the hate is mostly about the design.

So what if you can pick any car you want, based on how you like the design, and then just make it go electric. Wouldn’t that be awesome?

That’s exactly what the wizards at Stard Advanced Research and Development in Austria have done.

And it’s going to go full swing in 2020 racing at the World Rallycross Championship.

Imagine having the power to go 0-60 in 1.8 seconds. That’s faster than an F1 car!

The kit that they’ve developed has an all-wheel-drive drivetrain powered by 3 electric motors – 1 in front and 3 at the back, giving the car a total of 600hp, and a staggering 1100N-m of torque instantly available within 1 millisecond.

Even private teams can build their own electric rallycross cars based on the kit technology and the technical systems provided by Stard for a full and practical integration into these cars.

Check out the video. If you’re a true car guy, it’s going to be one of the best 4 minutes and 17 seconds of your life. You’ll thank me later.

Video © Lovecars

What I love about what Stard has done is that they manufactured it as a kit. Which means you can work with them for integrations with just about any car imaginable.

I bet if this becomes highly successful, and I’m pretty sure it will, that this will become an option for people who modify cars.

Ever thought about putting an LS V8 into your Mazda RX-7 track/drift car or Nissan Patrol offroader? Why not go electric!

Thinking about getting a Cybertruck but the look hasn’t grown on you? Turn your Toyota into an Electric Vehicle!

Driving your Supra and this Ferrari guy mocks you? Send all 1100N-m of torque into the pavement and leave him in your wake. You’ll just have to get used to not having those turbo blow-off valve sounds. LOL!

So tell me, what’s your favorite car, and if you had the chance, would you make it go electric?

The post Turn Your Car into a 600hp Electric Vehicle in 2020 appeared first on Robbie Agustin.

It’s #transformationtuesdays where I share use cases and success stories that you can learn from, and replicate too!

If your dream is to become a Data Scientist, and you’re wondering what you need to do, and what it would take to break into one of the most sought-after jobs of the future, then this article is for you.

Today, I’m sharing the success story of Jamen Long, and how he broke into his career as a Data Scientist.

And like you, he started with zero experience.

If you’re interested to find out exactly what Jamen did, then read on.

Don’t get me wrong. He did have work experience, though his background was quite different.

But the thing is, he really wanted to take the leap and jump into a career as a Data Scientist.

He knew he needed to learn stuff like R, Python, and SQL to get started. He did some research and took some online courses, but he didn’t really feel like they were allowing him to progress logically with a step-by-step path.

That is, until he found out about DataCamp.

From then on, his focus and direction became clear. And even starting out as a beginner, he was able to make steady, meaningful progress.

“DataCamp was essential for getting my job, which was my break into data science.” – Jamen Long

DataCamp is brilliant in such a way that you can learn it today, then apply it tomorrow.

In fact, Jamen is now a Data Scientist at Westfield Labs in San Francisco, thanks to the fact that he was able to leverage on DataCamp to acquire the skills needed to break into this career.

I’m sure you’re very curious right now, especially if you read up to this point. So I won’t delay you any further. You can read the full story by clicking on the link below.

Jamen Long’s Data Scientist Success Story

I hope you found value and inapiration in this success story.

Now if you’re interested in building a career as a Data Scientist, then here’s how you can take action:

• Visit DataCamp to start learning Data Science today! Heck, you can even start learning for free! Learn it today, apply it tomorrow.
• If you haven’t yet, download a copy of my book for FREE (for a limited time only) – The Business Optimization Blueprint, and learn how you can improve your business process and take it to the next level. I talk about practical uses of Data Science in business in Chapter 3.

The post How Jamen went from Zero Experience to becoming a Data Scientist (Success Story) appeared first on Robbie Agustin.

It’s #transformationtuesdays and our topic of the day is Data Science, and how Facebook uses it to do the following:

1. Facial Recognition
2. Textual Analysis
3. Targeted Advertising

No, they didn’t use some magic beans then all of a sudden they became an internet giant.

They uses Data Science extensively to make these things possible.

If you want to find out how, then read on.

I won’t add any more fluff. If you’ve read this far, I know you want to get straight to the point.

So without further ado, here are the use cases:

• Data Science at Facebook
• How Facebook Uses Data Analytics To Understand Your Posts And Recognize Your Face
• How Data Science at Facebook brings the world closer together

I hope you found this to be of value.

If you want to learn how to become one, some of the best learning tracks I’ve seen are the Data Science learning paths from DataCamp.

You can start learning today even for free. Check it out! Click the link below.

Data Science at DataCamp

The post How Facebook uses Data Science for Social Analytics (Use Case) appeared first on Robbie Agustin.