Welcome back to our course - AI for everyone
In this video we are going to see about, What is Machine learning? and What it can and cannot do?
Machine learning is an advancement of traditional AI technique. In traditional AI, the machine can do only the pre-defined things based on the given algorithm. But in machine learning technology, the machine will work algorithmically and also had the ability to learn new things on its own from the experience. That is, it has the ability to learn without being explicitly programmed.
Okay, now let’s see “what it can and cannot do?”
Machine learning can do the simple tasks when lot of data available related to that concept. That is, it requires enough data to do the task. Okay!!
But, machine learning cannot do the complex tasks when small amount of data available related to the concept. That is, it couldn’t do the task if enough data is not available.
Let’s see it with an example, consider the self-driving car. In that, AI can identify the nearby vehicles and other objects and know exactly where it is located. Then it will take decisions according to that. Because it collects enough data about the nearby objects or vehicles through it’s camera/sensors and all. So it will be really helpful for the self-driving car to go ahead with that enough data and take decisions accordingly.
But at the same time it cannot identify when the human is trying to gesture at your car, whether to ask your car to stop or to indicate that they want to take left/right and so on. Here the actual problem is that the number of ways people gesture at you is very large. So it’s difficult to collect enough data from thousands of different people.
As there is no enough data available obviously it will be difficult for the machine to take decisions to achieve the task.
In fact, even people will also find difficult to figuring out different gestures sometimes. Therefore, it’s actually a complicated concept for the machine to identify gesture and understand what exactly that person tries to ask. Therefore, it couldn’t respond to the human gesture as it doesn’t have enough data to understand the situation.
Okay!! That’s what about the machine learning can and cannot do.
Let’s recap.
Machine Learning or ML is an advancement of traditional AI technique as it has the ability to learn from its experience.
ML can do the tasks successfully when there is enough data available related to the concept. But ML can’t do the tasks when the enough data is not available.
In the example of self-driving car, we can able to collect enough data related to the nearby objects. Therefore, the machine can also identify the nearby vehicles or any other objects and make decisions according to that.
But it’s not possible to collect enough data about different human gesture. Therefore without enough data, the machine cannot make decisions when human gesture at your car.
Therefore, it can do the tasks when enough data available but it can’t do the task when enough data is not available.
The conclusion is the machine learning technology can be helpful for us to some extent still it couldn’t explicitly replace the human beings.
I hope you understand “What machine learning can and cannot do?”.
