3 Disciplines of Data Science

The objective of this post is to provide an overview of the 3 disciplines of data science and the significance of them. I hope to enable the reader in making a thoughtful decision about the direction of their career after reading this article. Most of the knowledge I have acquired is by working with different people in the industry, reading books and trying to segment the field for easy retention. Although, there is an overlap in how each discipline interact with each other but a lot to be gained by specializing in one of the discipline before branching off to other disciplines. I hope that the reader will be able to get a rough idea on which direction they can target to further their careers.

Data Science is am umbrella term which is used to describe activities that are making data useful, such as machine learning, statistics and analytics. To identify which tool to use to solve a problem in hand is to ask:

How many decisions would you like to make before you begin to make them?

1. Statistics

If you want to make few but important decision under uncertainty, then you would go for statistics. Few examples to consider and think about:

• What is the fair value of a 6-sided dice if you are given £ amount equivalent to the face of the dice? (Application: Market-Making Trading)
• What is the risk of lending £1m to a large corporate client? (Application: Commercial Banking Risk Management)
• You would like to know whether a control to monitor rogue trading is effective? (Application: Global Markets Compliance/Conduct Risk)

2. Machine Learning

If you want to make many, many decisions under uncertainty (Another way of describing automation), this is Machine Learning and Artificial Intelligence. Few examples to consider and think about:

• Whether stock price will go up or down in next 5 mins? (Making this decision every 10 seconds) (Application: Quantitative hedge fund trader)
• Whether a particular transaction is fraudulent and making this decision for all the transactions periodically? (Application: Wholesale Bank Compliance)
• Whether a particular field in a dataset can be inferred if unavailable? (Application: Data quality and cleaning almost can occur in every business/function)

Book recommendation: A must read and my current favorite book on the topic is The Society of Mind by Marvin Minsky. The link to the book can be found here. Although the book is not directly related to data science, it is a must read for anyone interested in AI or just generally improving their intelligence. Minsky was an American cognitive and computer scientist concerned largely with research of artificial intelligence (AI), co-founder of the Massachusetts Institute of Technology’s AI laboratory, and author of several texts concerning AI and philosophy. Check out his Wikipedia page for more info about his work. The book acts as a great framework to understand how minds work and how intelligence emerges from nonintelligence parts.

3. Analytics

If you are unsure on how many decisions you would have to make and you are just looking for some clues or inspiration based on the data. You are just figuring out the unknown unknowns, this is Data Analytics. Few examples to consider and think about:

• Whether company A should merge with company B? (Application: Investment Banking, Yes! A small part of Investment banking work can be considered a subset of data analytics)
• How can a large pharma increase its profits in next 3 years? (Application: Management consultant/CEO/CFO. Increasing profits is one of the most common problem encountered at that level)
• Which is the best Borough to buy a house in London for a family with young kids (primary school age)? (Application: Parenting)

Closing thoughts

Before jumping into deciding which field you want to specialize in, I would recommend to note that the excellence of statistician comes from rigor. They are very careful to protect decision markets coming into wrong conclusions. Machine Learning and AI engineers are more concerned with performance and work to improve efficiency of their program and minimizing the error rates. The excellence of an analyst comes from speed. How quickly can an analyst surf through vast amounts of data to explore trends, discover the hidden gems and interpret the results to bring in front of decision-makers.

If you find this article useful, I would be grateful if you can share the link with your friends, social media or emails. I would be even more grateful if you can comment with your thoughts, questions or feedback.

Thank you for your time and attention in reading my post.

Junaid Arshad

---

Comments

---