Machine Learning has become an attractive field for a variety of industries. Ever since the rapid growth available computers, machine learning has found use in almost all industries- biology, cybersecurity, finance, engineering, etc. But what is Machine Learning? And why is it working so well?

In layman terms, machine learning is taking the idea of how human learns and applying it to a machine. Through recursive training, one can learn to do a task with great accuracy and precision.

Machine learning works so well because it is able to adapt to the requirement of the problem and its ability to tackle highly non-linear problems very effectively.

In recent times, I have noted machine learning techniques are creeping into computational fluid dynamics. As of today, I am aware of researchers working on projects where they are trying to create closure models for turbulence using statistical inference and machine learning (see Duraisamy et. al.), use physics informed machine learning to solve partial differential equations (see Raissi et. al.), and using machine learning to solve fluid dynamics equations (see Brunton et. al.). Also, of interest is a review by Taira et. al. Data science techniques are proving to be an excellent tool to provide real-time control and could be the way forward towards discovering currently unknown phenomena (especially in turbulence).

It is an interesting time to be researching about fluids. Not only are we exploring new boundaries of fluid mechanics, but we are also re-learning what we know about fluid mechanics. All of this is happening with the use of data and the large amount of computational power at our disposal. The future is going to be based on data science and it is prudent to learn techniques that will be invaluable in the future.