The Leapfrog of Machine Learning


Machine Learning is an emerging branch in computer Science. The term is not new. It was being used for many years.

The truth is that the theoretical basis of this science did not change much. But its practical implementation have started to boost just recently.

With new technologies that made it more available, Machine learning starts to become some kind of “commodity” in the software industry and we see more and more companies adapting it in their products.

There are different reasons why we see today more Machine Learning implementations than we saw in the past:

1. New available Open Source software solutions

We see in the last few years many new software implementations that made Machine Learning more available to developers. If in the past it was only the expertise of scholars from the universities that had PhD’s, then as of today we see many developers that started to implement Machine Learning solutions without all the academical background.

Some implementations include for example:

Cloud solutions

Several new cloud solutions for Machine learning appeared recently. The good news they bring is that you can implement Machine Learning algorithms via API’s and you do not really need a complex implementation.

To list some examples:

Big Data

This is the real driver: In this era of the Internet, many companies collect huge amounts of data and form “Big Data” storage for their internal use.

While having this Big Data, those commercial companies can start to utilise Machine Learning algorithms on this data to gain new essence and open the company portfolio to new horizons.

As Machine Learning is based on statistics, and statistics need large amount of data, the Big Data became a major force for implementations.

New openings

One amazing phenomenon that appears now is that Machine Learning penetrates any field where there is data, regardless of its commercial necessity.

Just browsing the datasets in the Kaggle site would reveal all different types of data that people are interested in.

To summarise, I’d like to add a nice example of how this new research field affect traditional studies where it can now bring new insights that may be surprising sometimes:

In this example, scholars of Bar Ilan University in Israel use Machine Learning to research on the age of books in the Bible: