What the heck is machine learning, and why is it everywhere these days?

Unless you’ve been living under a rock, ignoring every big tech advance in the past decade, you’ve probably heard of machine learning. Whether it’s better fraud detection and prevention, the handy online recommendations made by Netflix and Amazon, revolutionary facial recognition technology, or futuristic self-driving cars, machine learning is powering the current artificial intelligence revolution. But what is it exactly? Here’s a handy beginner’s guide.

What is machine learning, and why does it matter?

Machine learning is an approach to artificial intelligence that’s focused on making machines which can learn without being explicitly programmed. Learning is a profoundly important part of what makes us human. If we’re going to build AI that can carry out tasks with human-like intelligence, we therefore need to make machines that can learn for themselves, based on their past experiences.

This is different to the classical symbolic approach to AI, in which programmers create step-by-step rules for machines to follow, rather than allowing them to discover insights for themselves. While machine learning still involves this classical style of programming, it combines those basic rules with knowledge that computers are able to gather on their own to grow smarter.

Oh, and there’s a whole lot of statistics in there as well. Today, machine learning’s massive success has led to it becoming the most dominant subset of AI that is practiced around the world.

Can you give me a basic example of machine learning in action?

Absolutely. Machine learning can achieve some pretty impressive feats in AI (think self-driving cars or teaching robots to autonomously interact with the world around them), but it’s also responsible for simpler, but still incredibly useful applications.

One good illustration of machine learning in action is the so-called “spam” filter that your email system most likely uses to distinguish between useful emails and unsolicited junk mail. To do this, such filters will include rules entered by the programmer, to which it can add numbers that — when added up — will give a good indication of whether or not the software thinks the email is good to show you.

The problem is that rules are subjective. A rule that filters out emails with a low ratio of image to text isn’t so useful if you’re a graphic designer, who is likely to receive lots of useful emails that meet these parameters. As a result, machine learning allows the software to adapt to each user based on his or her own requirements. When the system flags some emails as spam, the user’s response to these emails (either reading or deleting them) will help train the AI agent to better deal with this kind of email in the future.

It’s simplistic compared to how we learn as humans, but it nonetheless achieves the result of creating an algorithm that improves its performance the more knowledge it receives.

I’ve heard of data mining. Is that the same thing?

Not quite. There are a lot of statistical tools involved in machine learning, and a good knowledge of math is going to help you as much on a machine learning course as speaking English will help you on an English literature course.

There’s definitely some crossover between the two fields, but the main distinction is that data mining is about drilling down into a dataset to find information. Machine learning is about using data to work out how to predict future outcomes, or to train a machine to perform a certain task.

One way we’ve heard it explained is that data mining is finding a list of dance centers in Portland; machine learning is learning how to dance.

Are there different types of machine learning?

You bet! The major way of dividing up machine learning is to focus on how the machine learns. There are four main approaches: supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning.

Supervised learning involves training data in which there is a desired output. (This is the category our spam filter algorithm falls into.) Next up is unsupervised learning, in which training data doesn’t have clear outputs. Then there’s semi-supervised learning, in which there are few desired outputs.

The other type of machine learning — which is increasingly popular these days — is reinforcement learning. This ambitious approach to machine learning involves rewarding an artificial agent based on what it does. Reinforcement learning can help machines achieve feats like figuring out how to play video games through trial-and-error, based on working out what increases its score.

Where do neural networks come into all of this?

Ah, yes: neural networks. If you’ve followed AI at all over the past decade (or read our handy recent explainer article) you’ve almost certainly come across these brain-inspired machine learning tools. Deep learning neural networks are a big part of machine learning today, but they’re not the only part.

Because the field is all about figuring out how best to fit data into models that can be utilized by people, a crucial step for machine learners is working out how best to represent knowledge when they set out to solve a problem. Neural networks are one example of how this might be achieved. Others include support vector machines, decision trees, genetic algorithms, and more.

Which programming languages to machine learners use?

Like the question above, there’s no one answer to this. Machine learning is a big field and, with so much ground to cover, there’s no one language that does absolutely everything.

Due to its simplicity, and the availability of deep learning libraries such as TensorFlow and PyTorch, Python is currently the number one language. If you’re thinking about delving into machine learning for the first time, it’s also one of the most accessible languages — and there are loads of online resources available.

Java is a good option, too, and comes with a great community of its own, while C++ and R are also worth checking out.

Is machine learning the perfect solution to all our AI problems?

You can probably guess where we’re going with this. No, machine learning isn’t infallible. Algorithms can still be subject to human biases, and the rule of “garbage in, garbage out” holds as true here as it does to any other data-driven field.

There are also questions about transparency, particularly when you’re dealing with the kind of “black boxes” that are an essential part of neural networks.

But as a tool that’s helping to revolutionize technology as we know it, and making AI available to the masses? You bet that it’s a great tool!

Editors’ Recommendations

• What is an artificial neural network? Here’s everything you need to know
• A history of artificial intelligence in 10 landmarks
• A.I. creates some of the most realistic computer-generated images of people yet
• New algorithm helps turn low-resolution images into detailed photos, ‘CSI’-style
• That’s ‘Professor Bot’ to you! How AI is changing education

What the heck is machine learning, and why is it everywhere these days?

Unless you’ve been living under a rock, ignoring every big tech advance in the past decade, you’ve probably heard of machine learning. Whether it’s better fraud detection and prevention, the handy online recommendations made by Netflix and Amazon, revolutionary facial recognition technology, or futuristic self-driving cars, machine learning is powering the current artificial intelligence revolution. But what is it exactly? Here’s a handy beginner’s guide.

What is machine learning, and why does it matter?

Machine learning is an approach to artificial intelligence that’s focused on making machines which can learn without being explicitly programmed. Learning is a profoundly important part of what makes us human. If we’re going to build AI that can carry out tasks with human-like intelligence, we therefore need to make machines that can learn for themselves, based on their past experiences.

This is different to the classical symbolic approach to AI, in which programmers create step-by-step rules for machines to follow, rather than allowing them to discover insights for themselves. While machine learning still involves this classical style of programming, it combines those basic rules with knowledge that computers are able to gather on their own to grow smarter.

Oh, and there’s a whole lot of statistics in there as well. Today, machine learning’s massive success has led to it becoming the most dominant subset of AI that is practiced around the world.

Can you give me a basic example of machine learning in action?

Absolutely. Machine learning can achieve some pretty impressive feats in AI (think self-driving cars or teaching robots to autonomously interact with the world around them), but it’s also responsible for simpler, but still incredibly useful applications.

One good illustration of machine learning in action is the so-called “spam” filter that your email system most likely uses to distinguish between useful emails and unsolicited junk mail. To do this, such filters will include rules entered by the programmer, to which it can add numbers that — when added up — will give a good indication of whether or not the software thinks the email is good to show you.

The problem is that rules are subjective. A rule that filters out emails with a low ratio of image to text isn’t so useful if you’re a graphic designer, who is likely to receive lots of useful emails that meet these parameters. As a result, machine learning allows the software to adapt to each user based on his or her own requirements. When the system flags some emails as spam, the user’s response to these emails (either reading or deleting them) will help train the AI agent to better deal with this kind of email in the future.

It’s simplistic compared to how we learn as humans, but it nonetheless achieves the result of creating an algorithm that improves its performance the more knowledge it receives.

I’ve heard of data mining. Is that the same thing?

Not quite. There are a lot of statistical tools involved in machine learning, and a good knowledge of math is going to help you as much on a machine learning course as speaking English will help you on an English literature course.

There’s definitely some crossover between the two fields, but the main distinction is that data mining is about drilling down into a dataset to find information. Machine learning is about using data to work out how to predict future outcomes, or to train a machine to perform a certain task.

One way we’ve heard it explained is that data mining is finding a list of dance centers in Portland; machine learning is learning how to dance.

Are there different types of machine learning?

You bet! The major way of dividing up machine learning is to focus on how the machine learns. There are four main approaches: supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning.

Supervised learning involves training data in which there is a desired output. (This is the category our spam filter algorithm falls into.) Next up is unsupervised learning, in which training data doesn’t have clear outputs. Then there’s semi-supervised learning, in which there are few desired outputs.

The other type of machine learning — which is increasingly popular these days — is reinforcement learning. This ambitious approach to machine learning involves rewarding an artificial agent based on what it does. Reinforcement learning can help machines achieve feats like figuring out how to play video games through trial-and-error, based on working out what increases its score.

Where do neural networks come into all of this?

Ah, yes: neural networks. If you’ve followed AI at all over the past decade (or read our handy recent explainer article) you’ve almost certainly come across these brain-inspired machine learning tools. Deep learning neural networks are a big part of machine learning today, but they’re not the only part.

Because the field is all about figuring out how best to fit data into models that can be utilized by people, a crucial step for machine learners is working out how best to represent knowledge when they set out to solve a problem. Neural networks are one example of how this might be achieved. Others include support vector machines, decision trees, genetic algorithms, and more.

Which programming languages to machine learners use?

Like the question above, there’s no one answer to this. Machine learning is a big field and, with so much ground to cover, there’s no one language that does absolutely everything.

Due to its simplicity, and the availability of deep learning libraries such as TensorFlow and PyTorch, Python is currently the number one language. If you’re thinking about delving into machine learning for the first time, it’s also one of the most accessible languages — and there are loads of online resources available.

Java is a good option, too, and comes with a great community of its own, while C++ and R are also worth checking out.

Is machine learning the perfect solution to all our AI problems?

You can probably guess where we’re going with this. No, machine learning isn’t infallible. Algorithms can still be subject to human biases, and the rule of “garbage in, garbage out” holds as true here as it does to any other data-driven field.

There are also questions about transparency, particularly when you’re dealing with the kind of “black boxes” that are an essential part of neural networks.

But as a tool that’s helping to revolutionize technology as we know it, and making AI available to the masses? You bet that it’s a great tool!

Editors’ Recommendations

• What is an artificial neural network? Here’s everything you need to know
• A history of artificial intelligence in 10 landmarks
• A.I. creates some of the most realistic computer-generated images of people yet
• New algorithm helps turn low-resolution images into detailed photos, ‘CSI’-style
• That’s ‘Professor Bot’ to you! How AI is changing education

Visualize thousands of distant worlds in the Periodic Table of Exoplanets

The first extrasolar planets (exoplanets) were discovered in 1992 by Arecibo Observatory. Orbiting around a fast-rotating neutron star named the Lich Pulsar, they were three small planets named Draugr, Poltergeist, and Phobetor. The first planet around a star similar to our sun was discovered in 1995, and now we know of more than 3,700 planets outside of our solar system.

To categorize all these new discoveries, astronomer Abel Méndez‏ from the Planetary Habitability Laboratory has created a Periodic Table of Exoplanets with 18 categories based on size, temperature, and composition.

On the left side, the planets are slotted into one of three zones, roughly based on how water would exist on the surface as vapor, liquid, or ice. The warm or “habitable” zone exoplanets hold the most interest for researchers, as those are the locations most likely to support life. “Unfortunately, we don’t know yet if they also have the right amount of water (e.g. oceans) or the right atmosphere for life too,” Méndez explained to Gizmodo‏.

Across the top, the planets are split between rocky “Terran” and gaseous “Gas Giants.” Those two designations are further subdivided by size, ranging from Miniterrans (like our moon) to Jovians (such as Jupiter or Saturn).

The numbers show how many exoplanets of each type have been discovered, as well as the percentage of the total. You can see that 21 Warm Terrans were discovered that are the most similar to our Earth so far, or about 0.6% of the total.

Along the upper right, the numbers indicate how many stellar systems we’ve discovered, and how may exoplanets each one contains. For example, we’ve 16 systems have been uncovered that contain seven planets. Most stellar systems that astronomers have found contain at least one planet.

For comparison, the Solar System Periodic Table categorizes the 82 planets, dwarf-planets, and moons in our own system whizzing around the sun or orbiting various planets.

The majority of the discoveries are hot planets, but that’s likely because they’re much easier to detect, and thousands more are out there awaiting discovery. “I’m overwhelmed by the number and diversity of planets in the stars around us,” Méndez‏ said. “So many places to explore in our own Solar System, but much more is waiting for us beyond.”

Editors’ Recommendations

• This exoplanet telescope could help astronomers discover alien life
• Prepare for liftoff! 17 upcoming space missions worth getting excited about
• The 10 best exoplanets we’ve discovered so far, ranked
• Explore our solar system with Google’s new maps of Pluto, Venus, and more
• RIP, Cassini: 25 spellbinding Saturn images from NASA’s self-sacrificing probe

Verizon and NFL move forward on deal to expand streaming rights

Bloomberg has reported that Verizon, the largest wireless carrier in the U.S., is close to signing a new deal with the NFL that would grant it digital streaming rights.

Verizon’s current deal with the NFL allows the carrier to live-stream games to devices which have seven-inch screens or smaller, but this new agreement would allow Verizon to stream football games to any device, including TVs.

While this deal will expand Verizon’s streaming service, it will also cost it the exclusivity the previous deal offered. Under Verizon’s current agreement, it is the only company allowed to stream NFL content to smartphones. However, this new contract would allow competitors such as Sling and other mobile carriers the right to stream games through their smartphone apps.

Currently, the NFL has spread its content among multiple parties. Verizon currently has the right to stream Sunday day games in a team’s home market, in addition to Thursday, Sunday, and Monday night games. DirectTV has the rights to out-of-market Sunday day games, including on mobile devices.

For the past year or so, Verizon has been attempting to enter the already-crowded streaming market, but has failed to do so. One of the largest hurdles has been obtaining the rights to content. Rather than attempting to purchase media outlets or create its own content as Netflix does, Verizon is hoping deals such as this one will be enough to entice customers and advertisers.

Whether or not these deals will be enough to bring in customers remains to be seen. In a world with Netflix, YouTube TV, Hulu, Amazon Prime Video, Disney, and others, it may be difficult for any one company to make a dent in this market.

The NFL is likely hoping this deal will bring in new audiences as the league is currently in a bit of a slump due to declining ratings from traditional TV audiences. While cord-cutting is one of the reasons for the decline in NFL ratings, it is not the only one. Experts have cited numerous reasons, including the rising awareness of concussions, the poor performance of teams in major markets, and the recent protests by Colin Kaepernick and other players.

Editors’ Recommendations

• Disney is starting its own streaming service, pulling content from Netflix
• Nintendo rule on live-streams highlights contentious relationship with YouTube
• Hulu cuts the cost of its cheapest tier as Netflix prices rise
• Sprint wants to pull you in with free Hulu under its unlimited data plan
• Xbox One gets love from Amazon as it rolls out Prime video app to more countries

Master Microsoft Office with this training bundle for just $25

When it comes to office tools, Microsoft Office is the leading suite thanks to a wide variety of tools and apps contained within. Word or PowerPoint are pretty easy to get started with, but not Excel. There’s a steep learning curve involved there that isn’t easy to master.

In order to become proficient in all that Microsoft Office contains, it’s not a bad idea to enroll in a specialized course. Unfortunately, these courses are often very expensive, and you might not have time right now for anything extra.

Android Central Digital Offers has a deal on a lifetime subscription to eLearnExcel. This bundle, which you can access at any time forever, is normally about $1,000. Typically, you can get it for just $69 from iMore Digital Offers —That’s 95 percent off the regular price — but they’ve lowered it even further to just $25!

The bundle includes everything you need to know about using Excel. In all, there are eight courses and more than 280 lessons available.

If you’re ready to take your Microsoft Office skills to a pro level — you can even earn a CPD-certified master diploma in Excel — this is the bundle for you. Don’t wait too long; this price doesn’t last forever.

See at Android Central Digital Offers

Cheer yourself up in this weekend’s comments thread

The days are officially too short and too cold for many of us, so spend your time indoors talking to friends.

In a few days, the holiday season officially begins in the U.S. That means turkey dinners, Black Friday shopping and the end of Daylight Saving Time makes it dark before you get home from work. Winter time is fun for some things, but cold and dark is generally not a good for happiness.

While you’re mentally preparing for a family dinner complete with politics and religion and every other awkward thing being discussed over turkey and cranberry sauce, remember you have at least one place where nobody cares about any of it and you can relax and have a little fun. This place, right here.

We plan to take a friend to see some history this weekend. She’s never had time to see the sights whenever she’s visited my little section of the world before, so anything built in the 1700’s is fair game. Yeah, that’s not old if you live across the pond but for this part of the world it is. Living a stone’s throw from places like Harper’s Ferry means plenty of things to see.

Besides, I like taking photos of all the old stuff like the spooky church above.

I’ll have a nice Thanksgiving, and hope each of you can do the same. But that’s next week, so let’s hear what everyone is up to this weekend!

O2 bundling extra data and free Xbox Live with the OnePlus 5T

British customers can snag some decent extras with their OnePlus 5T thanks to O2.

As has been the case in the past, O2 is the carrier partner for the latest phone from OnePlus and it’s making a pretty big deal about its launch. There’ll be a handful of pop-up shops opening on November 21 for folks to go down and buy, as well as a bundle that includes extra data and free Xbox Live Gold when you buy on a new contract.

The information is being sent out by OnePlus to previous customers and it dropped in our inbox this morning detailing a pretty nifty bundle. O2 will be offering 50GB of data per month for the price of 10GB as well as a free 12-month subscription to Xbox Live Gold, because, reasons. The only catch is that it’s a limited time offer and you have to pre-order or buy the OnePlus 5T before December 6 to claim the Xbox Live offer.

The 50GB tariff will cost £49 a month with a £9.99 upfront payment for the 64GB model and £52 a month for the 128GB.

The pop-up shops will be open from 2 pm on November 21 and will be the first places in the UK folks can get their hands on the new phone. And maybe some swag.

These Pop-Ups will be open at O2 stores at London 134 Oxford Street as well as London Westfield Stratford, Manchester Arndale and, for the first time, also in Belfast at O2 Castle Lane.

O2 will also be selling off-contract but with a £15 charge for pay-as-you-go credit, with both storage sizes available for £449 for 64GB and £499 for 128GB. They’ll probably not be available at these pop-up stores for long and lines are to be expected so if you’re keen, get yourself down to your nearest nice and early.

See at O2

OnePlus 5T and OnePlus 5

• OnePlus 5T hands-on preview: Relentless iteration
• OnePlus 5T specs
• All of the latest OnePlus 5T news
• Join the discussion in the forums

OnePlus
Amazon

Recommended Reading: The church of AI

Inside the First Church of Artificial Intelligence
Mark Harris,
Wired

You may know Anthony Levandowski from being at the center of Waymo’s lawsuit against Uber, but he’s also the “Dean” or leader of a new religion of artificial intelligence. Wired takes a look at Way of the Future’s doctrine, Levandowski’s role and the quest to create the divine AI.

iOttie Easy One Touch 4 Qi Wireless Car Mount review: A driving essential

Mount and charge your phone wirelessly with the latest adjustable car mount from iOttie!

The iOTTIE Easy One Touch 4 Qi Wireless Car Mount is quite the mouthful for a product name, but it totally lives up to it. It’s stupid easy to use and will hold your phone in place and charge it wirelessly with not much more than a quick tap. Better yet, the mount is infinitely adjustable so you can mount your phone exactly as you’d like your car.

Earlier this year I took a look at the iOttie iTap Wireless Charging Mount which I found to work really well, but many commenters didn’t like the concept of attaching a metal plate to the back of their phone. That’s a fair point, which is why iOttie also offers a more traditional car mount that also supports wireless charging without magnets.

Given that this is the fourth version of this car mount, it’s safe to say that iOttie has got this phone mounting business down to a science. The Easy One Touch 4 can be oriented any way you like your car. The suction cup mount that’s installed on the mount out of the box is ideal for smooth surfaces such as a windshield — or even the smooth your laptop as I discovered by accident. However, since a majority of states and provinces have imposed a ban on windshield mounts that can obstruct a driver’s vision, iOttie also includes an adhesive pad to attach the mount to the textured surface of your dashboard.

iOttie sure seems to have got mounting phones in cars down to a science.

Once you’ve mounted the Easy One Touch as you like in your car, using it on a daily basis is as easy and convenient as the name implies. You squeeze the release tabs on the sides of the mount to expand the arms that hold your phone in place, then you simply need to press your phone against the One Touch trigger button on the backplate and the arms automatically snap back in and close securely around your phone. The mount works universally with the biggest phones out there and should accommodate even the bulkiest cases. Once locked in, you’re able to easily rotate from portrait to landscape mode or use the telescopic arm to perfectly position your phone so you can follow driving instructions or skip a track in Spotify quickly and safely.

It really is a great car mount — and that’s without even factoring in its Qi wireless charging capabilities. iOttie has includes a 12V car adapter and a Micro-USB cable for delivering power to the mount with a LED indicator on the top left corner of the mount showing when your phone is properly charging. The Qi coil on most phones is generally found in the center of the device, and iOttie has set the mounts coil so that it’s compatible with the latest Samsung phones such as the Galaxy S8. You can, however, also adjust the mounting foot lower if you find your phone isn’t lining up properly due to a bumper or case.

Don’t have a phone that supports wireless charging? You can still use a traditional charging cable thanks to the gap in the mounting foot, but you could also opt for the non-Qi charging Easy One Touch 4 which features the same great mount design at a better price without the wireless charging. I thought I might find the design to be a bit cumbersome or cluttered on my dashboard, but it’s ended up being a really minor trade-off when compared to how easy it is to use and adjust.

Just remember to always drive safe and keep your eyes on the road!

See at Amazon

These Android phones support wireles charging

Best Phones for Rooting and Modding

• Best overall
• Best flagship

Best overall

OnePlus 5T

See at OnePlus

OnePlus phones have been fan favorites among Android enthusiasts since the company was started because they’re cheap and run well. The 5T is no exception.

OnePlus makes it easy to unlock the bootloader and install alternative software, and the relatively low price will make for a large development community. A strange positive is that issues with the factory software mean that more people are interested in modding the OnePlus 5T, which makes a large community even larger.

If you are looking to spend some money on a phone just so you can hack away at it, the OnePlus 5T is worth looking at.

Bottom line: High-end specs and a budget price make the OnePlus 5T a great phone. Easy modifications and a large community also make it an excellent choice for rooting and modding.

One more thing: If you are holding onto an older OnePlus device, like the OnePlus 3 or 5, there are already plenty of custom ROMs available.

Why best

The OnePlus 5T is open and affordable.

OnePlus phones all share one common trait — they are easy to root and mod the software. The reason we think the OnePlus 5T is the best is that the low price also means the community is huge.

When you want to go with custom software the best thing you can have is a large group of people that act as a support channel. The OnePlus 5T’s low price also makes it a great phone to get if you’re looking for a second device to play around with while you use your primary phone to hold all your personal data without compromising its security.

Whether you’re new to phone modding or an old hat, we can’t help but recommend the OnePlus 5T as the best phone to do any of it.

The flagship option

Google Pixel 2

See at Verizon
See at Best Buy
See at Google Store
See ta Project Fi

While the standard Pixel 2 is the cheaper of the two options, it’s still not a phone anyone would consider inexpensive. But it’s still an excellent phone if you want to customize your software.

The Pixel 2 is packed with a Snapdragon 835 processor and 4GB of RAM. It’s available with up to 128GB of storage space. The thing simply flies, even without any tinkering. But because it’s a phone from Google, you’re able to unlock the software and do just about anything to it, all the while having a path back to “normal” in the form of factory software.

Bottom line: If you’re looking to mod the very best, the Google Pixel 2 is it.

One more thing: Unlocking the bootloader and installing other software doesn’t automatically void the warranty!.

Conclusion

You can root and mod almost every Android phone. We tend to focus on the ones you can’t because they are outliers. But being able to do it through an exploit or other sometimes difficult process isn’t ideal.

If you’re looking ahead and know you’ll want to change something on your next phone that requires custom software or root access, these phones are the best options.

Best overall

OnePlus 5T

See at OnePlus

OnePlus phones have been fan favorites among Android enthusiasts since the company was started because they’re cheap and run well. The 5T is no exception.

OnePlus makes it easy to unlock the bootloader and install alternative software, and the relatively low price will make for a large development community. A strange positive is that issues with the factory software mean that more people are interested in modding the OnePlus 5T, which makes a large community even larger.

If you are looking to spend some money on a phone just so you can hack away at it, the OnePlus 5T is worth looking at.

Bottom line: High-end specs and a budget price make the OnePlus 5T a great phone. Easy modifications and a large community also make it an excellent choice for rooting and modding.

One more thing: If you are holding onto an older OnePlus device, like the OnePlus 3 or 5, there are already plenty of custom ROMs available.

Updated November 2017: Moved OnePlus 5T to the top spot and added the Pixel 2 as the flagship option. Goodbye, Nexus 5X, we’ll miss you and your $200 price.