Misunderstanding data science and analytics

Students in Management institutes pan India are today queueing up for courses on analytics with a hope to make a dent in the data science domain. The common misunderstanding is that capability of writing code and using data visualisation tools is a sufficient qualification to be successful in data science. We need to think this through again. Any data science application, for that matter application of any management technique in business should be with a view to answering a simple question " Does my input solve a business problem". The answer should be satisfy ourselves first and then the business we are in. Given today's complex business scenario, analysis and prescription to a business problem can be broadly divided into two parts- The algorithm which models the business problem and the code which runs the algorithm on the given database. Algorithm is based on understanding of the business and evaluation techniques used to model the business. A substantial part of algorithm in today's "big data" world is rooted in a strong understanding of fundamental and advanced statistics and linear algebra. The "code" and "data visualisation" part is constantly evolving and we are fast approaching the "no code" age thus making expertise in writing code a "redundant" skill . AI take a bow!!!. AI might help with the algorithm formulation also but only if the model is fed into it (till such a time that it learns the business itself!!). Therefore students would do well to pay more attention to the algorithm development part if they wish to be relevant in this domain. Code will soon be a test of English language !!!