Programme

Registration and Refreshments

Chairman's Welcome

Chairman:
Andrew Dellbridge

Advanced Analytics and Data Mining

• The essential principles
• Which method for which purpose - the breadth and depth of the tool box
• Describing v predicting, classical methods vs. artificial intelligence
• Using multiple techniques in series and in parallel
• Important do’s and don’ts
• Where and how it worked well and where it failed to deliver and why
• Case Study: DMA Targeting - which techniques work well, and which to avoid

Speaker:
Barry Leventhal

Factor Analysis (including Principal Components Analysis)

• The principles. How it works. When to use. When not to use 
• Common factor analysis vs. principal components analysis
• Applying the method in practice: the stages to go through and the pitfalls to avoid
• Interpreting the results: factor loadings, factor rotation
• Selecting optimal factor solutions - scree plots 
• Distinguishing explanatory and confirmatory factor analysis
• Tips for best use
• Practical application of the insights obtained - case study

Speaker:
Dave Walter

Cluster Analysis for Customer Segmentation

• When to use Cluster Analysis - what it can deliver
• The main clustering techniques
• How they operate
• Principle outputs
• Pros and cons
• Developing a segmentation - best practice
• Maximising the benefits of segmentation and avoiding pitfalls
• Two case studies - one in the mobile telecoms sector and one in the public sector

Speaker:
Corrine Moy

Refreshments

Decision Trees

• The principles and principal features
• Decision Tree algorithms deconstructed: CHAID, Gini, C5.0, CART and others
• Use in data exploration and in model building
• Using Decision Trees in response modelling
• Understanding and interpreting the output
• Automatic versus manual model building and pruning
• Pitfalls, tips and best practice – using ING as a case study example

Speaker:
Tom Breur

Advanced Application of Regression Analysis

• The context of modelling in insurance and the typical model forms which are appropriate
• Multiple regression v logistic regression
• The stages in building and evaluating a regression model
• Tips, traps and pitfalls
• Understanding the importance of the ‘Case Deleted’ Deviance approach and if it can help to obtain improved models. When to use and when not

Speaker:
Tony Lovick

Advanced Data Visualisation

• What is ‘Advanced Data Visualisation’?
• Best practice principles - the do’s and don’ts
• The main techniques for visualising data
• Case study examples of visualisation applications:

           - Data quality analysis

           - Exploratory analysis of multidimensional data

           - Segmentation analysis including geodemographics

           - Social media analysis

           - Analysis of real-time spatially referenced data

• Overview of the visualisation tools available in the market place
• A glimpse into the future

Speaker:
Peter Furness

Lunch

Analysing the Social Networks of Customers

• Introduction to Social Network Analysis
• Business applications of SNA
• Extending SNA to Social Media Analysis
• Case study using a mobile phone operator

Speaker:
Judy Bayer

Web Analytics

• Web Analytics’ and ‘Marketing Analytics’: similarities and differences, partners or competitors?
• Differences in data sources between online and offline marketing and their implications
• What web analytics can deliver and the methods to use
• Integrating web analytics with analysis of offline customer behaviour
• Case study: MacDonald Hotels - showing how online is about combining web analytics with targeted email communication and qualitative data like usability to maximise improvements
• Case study: eGroup - showing how web analytics data can be used to spot online fraud

Speaker:
Gary Lee

Panel: Problems and Solutions in Realising the Benefits of Advanced Analytics

• The criteria that determine whether or not a business will successfully deploy advanced analytics and achieve its goals
• Pitfalls commonly encountered - how to overcome them
• Resolving common and less common cultural and organizational issues
• Solving technical problems and overcoming technological barriers - the solutions

Contributions from:

Gordon Farquharson