Suraj Aggarwal

gdgdgfhfhf fh hh hh

The time has come when computer, the so called electronic brains, can have the human brain like learning capabilities. The domains where the learning capability can be expertises in, are countless. Let it be finance forecasting, learning to play games, weather forecasting or the domain concerned to us i.e. driving cars autonomously. Making machines learn needs-

Existence of pattern

Inability to pin down the learning objective mathematically.

Lots and lots of data i.e. examples to be fed to the machine to learn from.

An attempt to make machines learn without above prerequisites-In an attempt to make machine learn without pattern will only let you that – there exists no pattern and the machine is not yielding relevant answers for the questions asked to it. Existence of the mathematical formula will let your machine learning attempt go in vain since the machine will only be able to approximate the correct mathematical formula thus giving a lesser performance measure than possible with the exact formula. Need not to mention, machine learning is not at all possible without the data, moreover you ought to have lots and lots of data. Machine learning has the capability to be the backbone of the all the working modules in AGV. Since the practical driving conditions are highly unpredictable; the power of machine learning can be rightly exploited; in controls, perception and planning; to train computers, much like humans, to drive cars even in the most robust and unfavourable conditions. With the experience computers only become more and more capable and learned to deal with the adverse driving conditions even more precisely than humans can.