Course Details

Print pageDownload as PDF
We Invite You to a Course in

Multivariate Data Analysis, General

Multivariate Analysis, General
Using the latest multivariate techniques, participants will learn how to interpret complex data quickly and confidently. Discover the secrets of overviewing data tables and also learn how to build robust predictive models that turn data into decisions.
The course is composed of lectures, demonstrations and computer exercises in software SIMCA-P+, based on real-life datasets.

After completing the course, participants will know how to:
• Apply PCA (Principal Component Analysis) to detect outliers, trends, patterns and to classify groups within complex data
• Relate multiple responses to multiple inputs using PLS (Partial Least Squares) modelling
• Interpret models to gain scientific insights
• Learn how to construct predictive models and apply them in practice. Use multivariate calibration to predict and improve quality
• Classification of raw materials and ingredients
• Analysing a multivariate problem.

Who should participate?
The course is intended for researchers, scientists and engineers from all sectors of industry and academia. No prior knowledge of statistics is assumed.

Course schedule
Lunch daily between 12:00 - 13:00
Day one
9:00 Introduction and presentation of the need for multivariate data analysis. Introduction to three kinds of problem: overview, classification, and quantification and prediction.
Principal Component Analysis (PCA) for over-views of data tables: variable scaling, geometrical interpretation, and model evaluation.
PCA examples: quality control of manufacturing, process control (SPC) and environmental monitoring.
Computer exercises followed by discussions.
17:00 End of lectures and exercises.
Day two
9:00 Partial Least Squares (PLS): prediction of responses Y from control parameters X.
Variable scaling, geometrical interpretation, algebraic solution and model evaluation.
Model diagnoses and validation of a PLS model.
PLS examples: process and quality control. 
Computer exercises followed by discussions.
17:00 End of lectures and exercises.
Day three
9:00 Multivariate calibration: prediction of chemical contents from spectroscopic data. Multivariate classification; two types of classification techniques will be discussed; SIMCA classification and PLS-DA (discriminant analysis). Exercises and discussion of participants’ own data. Course summary.
15:00 End of course.

Cost and conditions
Early bird registration: if you book at least 2 months prior to course date the price DISCOUNT is 17%. Also if you are more than three participants from one company booking the same course date you will receive an 8% DISCOUNT/person.
Course fee (+VAT) includes coffee, lunch and course documentation. An invoice will be sent and payment is required within 30 days of the invoice date. Course application is binding. Cancellations registered later than two weeks before the course start will not be refunded. Provided that Umetrics AB is notified, the registering company may substitute its participant(s).
If the customer cancels prior to 14 days of the course, 10% of the class fee will be applied to cover processing costs. The balance of class fees already paid to Umetrics may be credited towards a future course. The registering company may substitute its participant(s) provided that Umetrics is notified.
Umetrics holds courses based on a sufficient number of registrants. Therefore, Umetrics reserves the right to cancel the course 14 days prior to the course start date if the number of registrants is too low. Full refund will be made to these registrants. Alternately a 10% refund will be made to any registrant(s) enrolling in the next available course.

To register, please use the button "Begin registration" in this window frame, or send an email to Umetrics Academy.

COURSE INFORMATION

Course length: 3 days
Location: Switzerland, Basel
Start date: 2010-11-09
Price: 1200 EUR
Contact: umetrics.academy@umetrics.com

Please visit www.umetrics.com/training for more information and registration.

No prior knowledge of statistics is required, knowledge about in-house data base structure is beneficial.