Module Directory

The module covers two important aspects of managing data within any organisation. First, the fundamental concepts that underlie the techniques and algorithms deployed by data science teams to extract information from data. Second, the framework that aligns the understanding of three different types of teams that managers need to coordinate: business team, the technical/software development team, and the data science team.

While the module does not require a sophisticated mathematical background, the content of the module by definition being technical, we cannot exclude it. Nevertheless, the exposition involves significant data handling, software skills and associated data computations. Expect a hands-on and hands-full, experience. The focus, however, is on those conceptual aspects of data science with which managers increasingly need to be familiar.

The aims of this module are:

• To describe and explain the key principles employed in data science, and how they are applied in contemporary organisations.

Knowledge of the key principles employed in data science will:

• Equip managers with an awareness of, and the ability to critically assess alternative ways of envisioning the business problem from a data-centric perspective.


• Enable managers to select, from different data science techniques, those suitable and viable ones for the business problem at hand.


• Enable managers to distil insights from the results of the data mining techniques selected and identify those insights that can profitably be deployed.


• Enable managers to identify the need to iteratively address the business problem that was initially envisioned and finally to help decide when to stop the iteration cycles.

By the end of this module, students will be expected to be able to:

1. Demonstrate a comprehensive understanding of the organisational setting into which data science fits, including how to select, structure and retain data science teams, how to develop competitive advantage when employing data, and demonstrate awareness of tactical concepts for running data science projects.


2. Select and apply suitable lenses to identify the relevant data to invest in and the methods to acquire that data.


3. Explain data mining processes and understand how high-level data mining tasks need to be delegated.


4. Demonstrate a systematic understanding of the concepts underlying the vast array of algorithms for prospecting and mining the business-relevant knowledge from the vast data that is now increasingly flowing into organisations.

Teaching Programme
Indicative Lecture Programme

Monday
Session 1:
Chs. 1 & 2.
Introduction – Data Analytic Thinking.
Business Problems & Data Science Solutions.
Session 2:
Ch. 3.
Prediction, Entropy & Information Gain.
Segmentation & Strategy, Classification Trees.

Tuesday
Session 3:
Ch. 4.
Fitting a model to data (Estimation), OLS, Support Vector Machines.
Session 4:
Ch. 5.
Overfitting, Holdouts, Cross-Validation.

Thursday
Session 5:
Ch. 6.
Metric Spaces: Similarity, Neighbours, Clustering & Marketing.
Session 6:
Ch. 7.
Data Analytic Thinking: Expected Value, Type I & II errors, and Cost -Benefit Matrix.

Friday
Session 7:
Ch. 8.
Evaluating Performance of Models: ROC, AUC, Lift Curves.
Session 8:
Ch. 10.
Text Mining, Sparseness, TFIDF.

This module will be delivered via:

• Four days of teaching in one week.

In the teaching week, there are eight teaching sessions on every weekday except Wednesday. Teaching starts at 1000 each day and lasts till 1700, with breaks in between obviously. Overall, you should budget for 100 hours in total to the module. Of this total, the teaching comprises 24 hours while the remaining 76 hours are an estimate of the additional hours of your own time (typically spread over two weeks). The lectures and computation slots during the teaching week comprise 24 hours. The teaching is distributed over four weekdays in one week.

The lecture content is described in this outline. In general, you are expected to do some (by no means all) relevant reading and preparation before the associated lecture. There is one textbook for this module, and it is by Provost and Fawcett (PF). Provost teaches from this textbook on the MBA at Stern, NYU and the late Fawcett co-taught with him.

As you will have already come to realise, several modules on the MBA are inter-related. This module is related mainly to marketing and to operations and partly to strategy and to finance. I am sure you will notice the overlaps as we proceed. I would like you to engage with the issues and concepts associated with managing data science teams, a skill that will almost surely future proof your careers. This is my valid argument, even though there is not solid and sound evidence, my argument rests on the evidence that mere observation and press reports confirm the increasingly important role that artificial intelligence (AI) and the associated algorithms and techniques play in disrupting careers and organisations. While you should know enough detail concerning the quant tools and techniques, please do not get engrossed in the detail that the quant techniques and computing tools are meant to illustrate