If all those internet buzz words still blurry in your mind or you just heard everyone is murmuring around or even not yet clear-cut the differences. Well, you’re in the right place with no need to worry about it simply because we’re all in the same learning journey.


Machine Learning—ML

It’s always connoted with Artificial Intelligence. But what the heck is exactly about? It’s all about learning! In other word, teaching those adorable computers via a data loading process how to juggle various tasks like recognizing a human face then spotting a thief entering your home, self drive your car and park it in your designated lot, and a numberless of different uses in life.

This learning process is not an instructive classical programming work that goes step-by-step “if a=1 — c=2 — b=3 then put d=4”. Not at all! The developer doesn’t have to write any script to ask a machine to unlock the door for a stranger. The machine will run a neural networks process on its existing data base at first hand.

Then it should identify who’s standing at your door bell so to decide what to do next automatically! But this can be achieved upon a respective number of training or what is called data sets labeling in respective with data feed to the machine.

In this case, we’re talking about a supervised learning approach which is one of the three main learning types in ML. It basically teaches machines by examples! That’s why massive labeled data is quietly required so for a computer to succeed a particular task.
Labeled data is just a group of samples previously tagged with one or more labels. Think of a puzzle you’re trying to assemble its pieces you’ve seen already on a catalogue! Simultaneously, there are countless datasets used to train those machines or systems like the 8 Million labeled videos dataset from Youtube repository, a break tough research project intimated by Google AI.

The second machine learning typology is the unsupervised learning. In contrast to the first one, here the system doesn’t require any preload of labeled data sets. Thus, the learning process happens on the go while the machine is trying to spot similarities on a particular unlabeled existing data to be classified eventually into categories.

It’s like giving you a mix-up puzzle and asking you to figure it out! Google News algorithm for instance, which is guarded as Colonel Sanders’ secret recipe for fried chicken, collects and ranks enormous articles then it clusters them into news headlines. Ta-da! You’ve got your preferred first-in-line news.

The last methodology stands for the semi-supervised learning. This approach perches between the supervised and unsupervised leaning techniques. Meaning, the machine is fed with some labeled data and tremendous amount of unlabeled data to fulfill the training purpose.

The beauty if this process is it can give a birth to completely new data!Imagine for example, asking your super computer to create a new design of your Mercedes-Benz E-Class Cabriolet from existing designs data. WOW! I can imagine. But here is the trick, teaching machines in semi-supervised way will produced more labeled data sets to become then as effective as supervised learning.


Virtual Reality & Augmented Reality—VR & AR

Everyone of us remembers the famous humanoid movie Avatar released in 2009. It was the first one to introduce three-dimension-3D and virtual reality technology against its peers in Hollywood. In fact, the movie’s scriptment was completed in 1994 for a planned release in 1999 but according to James Cameron, director of movie, the technology needed a catchup with his vision at the time.

Today, VR technology is not anymore a science fiction or a fantasy movie to be watched in public cinema but a reality that is evolving very quickly across copious areas in our life. Yet, what is virtual reality or VR? It’s a broad term implied to a technology that attempts to create an authentic 3D image of any whatchamacallit you may think of in life. In technical word, VR is multi-sensory computer-generated experience aiming to simulate virtually an environment in full-fledge sensations.

To convert this definition into a real-life simulation, you need to simply buy a VR headset designed to run on a PC with VR software. Hocus-pocus! Welcome to VR world!

Does it sounds fair enough? If not, don’t scroll down unless you understand what differs AR from VR. The Augmented Reality or AR could be the most thrilling technology we’ve ever seen! That’s the reason why is little bit tough to define it. Compared to VR, AR technology is not easy accessible for somebody to try it personally may seems witless.

Anyhow, virtual reality tries to immerse us in a simulated real life situation whereas augmented reality does entirely the opposite. So, AR augments your real world by layering and overlaying it before your eyes. Then, don’t bewilder to see 3D-Tyrannysaurus dinosaur crossing the main street with pedestrians!

I know you have been exposed to this scene before, or introduced to the famous game Pokemon Go, played with Sony PlayStationVR, or if you may tried to use this cute puppy features of Snapchat on your smartphone. Well, this is how AR technology was initiated to us! Still, big tech companies are racing incessantly around the clock to develop their own smartphone-driven AR platforms like Apple’s ARKit or Google’s ARCore. Comparably, Microsoft HoloLens and Google Vuzix have been working to bring AR to us in eyeglass form.

Now, the question what are these technologies for? The beauty of its serious application would ameliorate various arenas and fields in our life. For example, VR is being used as an effective way of treatment in various healthcare applications such as autism or restoring low vision for elderly people. VR can teach medicine students by watching VR surgeries.

In construction, the $10 trillion industry is witnessing lot of astounding AR developments like OpenSpace technology, which is a mix reality of VR plus AR added to AI, enables project managers to walk through a project over its completeness stages like Google Street View. Another AR application developed by DAQRI uses smart helmets to deliver site-specific information to builders in real-time.

In Enterprise training, everybody knows pilots have been trained in their VR simulators for decades. Today, other heavy industries where the risk of injuries and costly mistakes are pretty much higher can benefit from VR simulations as apposed on job training!

In short, technology still has endless areas to solve real problems across a variety of industries that besiege this world.

(Visited 49 times, 1 visits today)