The best way to learn data science is to start with understanding data – as it forms more than 70% of the overall work involved in data science projects. Students need to understand data engineering i.e. how to treat and manipulate raw and unstructured data in a way so as to make it eligible to run statistical analysis on top of it.
Once this is done, the next step is to learn any scripting language in-depth preferably Python which can be used to apply statistical models on the cleaned data. The understanding of a scripting language would enable the student to implement the concepts learned with real-world data and see the outputs. Finally, once the student is comfortable with the basics of statistics and Python; he can then proceed towards more complicated statistical models & predictive analytics.