What is Data Science?

Data Science is not born out of nowhere, ever since the time we have the basic algebra, we start to apply the algorithm (1+1=2, oh yeah!), combined with our data (how many lands are there?), to generate important insights (which crops to grow next year). While life is much easier that time, isn't it?

Nowadays, with the paramount of data generated every day, Data Science gradually becomes one emerging study subject to leverage machine learning algorithms, utilize "big data",  and deliver advanced data products and meaningful business insights.

Python

"Life is short (You need Python)" -- Bruce Eckel

Python is very easy to use, has a great ecosystem, and data science friendly. Its numerous packages, like numpy, scipy, pandas, scikit-learn, theano, plotly, folium, jupyter notebook etc., provide great data science tools. Comparing with R (another popular statistical language), Python is well accepted in Engineer's world, so any data scientist want to get their code fast into production should learn Python.

Machine Learning

"All models are wrong, but some are useful" -- George Box

Machine learning is becoming a "buzz" word, a machine can learn everything! However, it is merely algorithms that can solve specific tasks, like solving regression, classification, clustering, or even play Go! Algorithms could be complicated or quite simple, but it is about how to identify and use the most appropriate one to solve the business problem.

Spark

"May the Spark be with you" -- A Spark Summit T-shirt

How to put the "big data" into computer memory to do the fast calculation? as if the data is not that big. Spark would be the answer. It leverages advanced in-memory technology to handle terabyte data to put them into action. With DataFrame/SQL, MLib, Graphframe and more all bundled together, it evolves into a must-have ecosystem for data scientist dealing with big data.

Business

Business insights and domain knowledge are very important to set the right direction on where to unveil the power data science provides. After all, data science is (just) a great tool to develop advanced products or understand complex business situations. Integrating data scientist with domain knowledge expert is always a good practice.

Some nice pictures :)