|
These courses are in the planning stage. We are aiming to deliver in 2009 and we would invite you to have input into the content and structure of these courses.
Please email your feedback to training@amadeus.co.uk or call 01993 878287 and ask to speak to our Technical Team about these courses.
TIME SERIES ANALYSIS/FORECASTING
Prerequisites
Familiarity with basic statistical concepts, such as hypothesis testing and regression, such as could be obtained by attending Amadeus’ business statistics course.
Objectives
This 2-day course demonstrates the use of new SAS techniques which bring modelling of time series within reach of regular business analysts, rather than specialized forecasters. The course will concentrate on these topics:
- Why time series are different
- Preparing data for time series analysis
- Analysing univariate time series using smoothing techniques
- Structural models for univariate time series
- Estimating the effect of covariates, leading indicators, interventions (campaigns)
- Assessing model fit
The course is largely based on Proc Forecast and UCM. Enterprise Guide will be used wherever possible, but some familiarity with SAS code is desirable.
At the end of the course, delegates will be able to:
- Create time series from transactional data
- Determine an appropriate analysis method, and fit a model
- Generate realistic forecasts, together with levels of confidence
- Quantify the effectiveness of factors that drive response, such as the effect of advertising or promotions on sales.
BUILDING SCORECARDS USING Base SAS and SAS/STAT
Prerequisites
Delegates will need to be familiar with programming using SAS.
Objectives
This 2-day course will show how SAS software can be used to build credit application scorecards. The course will cover the process from inception – the need for a new scorecard – to the point at which a new scorecard can be implemented. The following topics will be covered:
- Assembling the data for modelling
- Identifying the most informative variables to use in the scorecard
- Grouping continuous and discrete variables using Weights of Evidence
- Performing Reject Inference
- Building a preliminary scorecard
- Assessing the quality of the scorecard
- Shaping the scorecard
- Setting the cutoff
- Running scorecard report
At the end of the course delegates will be able to build their own scorecard using SAS software, for use in fields such as credit scoring, and understand the characteristics, strengths and weaknesses of their scorecard model.
TRACKING, MONITORING AND REVISING APPLICATION SCORECARDS
Prerequisites
Delegates require basic SAS programming skills and an understanding of the concepts and techniques of scorecard building, such as could be obtained by attending the Building Scorecards course
Objectives
This 2-day course will build on the material presented in the “Building Scorecards” course. The emphasis will be on ensuring that credit scorecards keep performing at the level attained at the modelling stage. The following topics will be covered:
- Assessment of in-use performance – discrimination and calibration, system stability reports
- Characteristics drift analysis
Vintage analysis
- Identification of groups of applicants showing scorecard bias
- Revising cutoffs
- Residual analysis (Marginal information)
- Simple and multiple realignment
At the end of the course delegates will be able to critically evaluate the operational performance of their scorecard, and take the steps necessary to identify whether and in what way their scorecard is underperforming, and take the steps necessary to recalibrate it. |