Advanced Certificate in Business Analytics for Data-Driven Decision Making (with Python)

Overview

A one-size-for-all approach to business no longer works. Segmenting and personalising based on data is one of the most powerful tools available to marketers and product managers today. Organisations should design experiences tailored to their users' individual behavioural, psychographic, demographic and characteristics to surpass competitors and get more customers.

Unsupervised machine learning algorithms such as K-Means clustering and hierarchical clustering are typically used by data scientists to develop customer segmentation models.

Participants will gain a better understanding of what conditions require segmentation and personalisation of customers, and how to apply computational business analytics to this process.

Topic/ Structure

• Introduction and Program Overview
• Mapping Types of Business Analytics Questions
• Customer Segmentation with Clustering
• Introduction to unsupervised machine learning with K-means and hierarchical clustering
• Data-Driven Customer Journey Mapping
• Market Basket Analysis
• Recommender Systems Part 1—Collaborative Filtering 
• Recommender Systems Part 2—Content-based Filtering

Overview

Predicting Customer Lifetime Value (CLV) is an important metric since it helps organisations understand their customer's lifespan, purchasing behaviour, and revenue from that customer. One of the best ways to predict CLV is with machine learning. This allows organisations in predicting customer behaviour, helping them to optimise marketing efforts, initiatives and budgets.

In this module, participants will learn how to answer business analytics questions about customer relationship management. Through supervised machine learning (ML) algorithms, they will learn how to predict CLV as well as translate ML performance metrics into business performance metrics. They will also learn recency, frequency, and monetary (RFM) analysis and survival analysis.

Topic/ Structure

• BG/ NBD (Beta Geometric Negative Binomial Distribution) and Gamma-Gamma Models
• RFM (Recency-Frequency-Monetary) and CLV (Customer Lifetime Value)
• Xgboost (Extreme Gradient Boosting) Algorithm and CLV 
• Predicting Customer Attrition with Supervise Machine Learning
• Classification Performance Metrics 
• Translating Performance Metrics into Business Metrics
• Review of Predicting Customer Attrition

Overview

In any organisation, strategic decision makers commonly face questions about the impact of a certain strategy on their target groups of people. This requires expertise in causal inference. It is notable, however, that causal inference is often missed in a Python course on computational business analytics.

Causal inference can be used to make information that can help improve user experience and generate business decisions by knowing its impact on the business.

In this module, participants will learn when business analytics require casual inference, how to execute in Python, and how to interpret the results. Additionally, they will also learn the logic of causation, fundamentals of A/ B testing and analysis, and econometric methods for analysing causality in naturally arising data. 

Topic/ Structure

• Causal Models, DAGs, and Threats to Validity
• Randomisation and Matching
• Difference-in-Differences 
• Regression Discontinuity
• Instrumental Variable
• Regression Discontinuity and Instrumental Variable Revisited
• Review of Regression Discontinuity and Instrumental Variable

Overview

Technology has become an asset in finance. With technology, the speed and frequency of financial transactions, as well as a large amount of data available, contribute to the increased focus of financial institutions on technology over the years, and this has led to technology becoming the main enabler in finance.

Python programming language has become increasingly popular for algorithmic trading, where financial instruments are traded using rules or algorithms with little or no human intervention.

In this module, participants will gain a deeper insight into how process-driven investing works, how Python can be used to extract historical stock data, as well as formulating based on historical data theory-driven and data-driven trading strategies.

Topic/ Structure

• Financial Data Acquisition/ Extraction and Wrangling
• Developing Theory-Driven Trading Strategies
• Backtesting Theory-Driven Trading Strategies
• Review of Development of Trading Strategy and Backtesting
• Developing Data-Driven Trading Strategies
• Backtesting Data-Driven Trading Strategies
• Developing and Backtesting Multiple Trading Strategies
• Review of Data-Driven Trading Strategies

Overview

Real-world business analytics questions inherently deal with text data. This data may come from customers, employee reviews, speeches and publications of organisation leaders, and articles in the media.

In this module, participants will learn how to deal with such text data, with a particular focus on customer and employee reviews. The module will provide lessons on pre-processing text data, natural language features, and feeding features developed from text mining into modelling processes. Participants will run sentiment analysis and execute unsupervised learning on text data with topic modelling to discover hidden clusters within them.

Topic/ Structure

• Types of Natural Language Features
• Sentiment Analysis
• Use Machine Learning algorithms for Text Classification 
• Topic Modeling
• Structural Topic Model (STM)
• Applying Topic Models to Employee Reviews
• Applying Topic Models to Customer Reviews

Overview

For its optimal performance and usage, business analytics should not end with the analysis itself or with mere reporting on documentation; it needs to be “deployed”.

Participants will be guided through the process of deploying their business analytics work into an interactive dashboard - the Python Dash application. Along the way, participants will create Dash apps building on their business analytics work executed in response to assessment questions from the previous modules. With Dash apps, users will be able to interact with analytics work written in Python, thereby converting complex operational analytics into business intelligence.

Topic/ Structure

• Interactive Dashboarding with Dash
• Customer Segmentation Dashboard
• Customer Lifetime Value Dashboard
• Customer Churn Dashboard
• Program Evaluation Dashboard
• Trading Strategy and Backtesting Dashboard
• Topic Modeling Dashboard