Simulator now detects hidden patterns in datasets

Edited by the Engineeringtalk editorial team Nov 12, 2001

Witness Miner is an innovative new tool within the Witness simulation and optimisation family which provides the ability to detect hidden patterns in datasets in a unique manner

Witness Miner is an innovative new tool within the Witness simulation and optimisation family which provides the ability to detect hidden patterns in datasets in a manner that is unique in the current data analysis market.

Today's organisations collect a large amount of operational data relating to all kinds of activities.

If properly analysed, this data can have a significant effect on a company's performance and profitability.

Witness Miner provides both a useful tool and the project framework for such investigation.

Witness Miner provides knowledge from raw data through a powerful process for analysis and rule evaluation, whilst also offering easy data modelling and the facility for high quality reporting.

Witness Miner, with a friendly and modern interface allowing for easy data model construction and analysis, allows an exploration of data to determine fundamental relationships that affect business.

This new software package is designed to the latest software standards and makes the rigorous analysis simple to set up and easy to understand.

Effective business demands careful examination of data sources and analysis of simulation results.

Data such as breakdown data, people flows and tables of simulation results may include many hidden patterns which need to be understood to construct accurate models and take wide ranging decisions.

Lanner has already engaged in a number of successful applications of such data mining techniques.

In the medical sector the company has used data mining to analyse a database of clinical records for diabetes patients.

This identified patterns in the rules found and described a scenario not generally known by the medical community.

In a Customer Relationship Management application, for a major breakdown and recovery service provider.

Lanner's data mining tools flagged-up cross selling opportunities to identify customers with the best chance to sell new financial products.

For warranty data analysis.

Lanner used data mining in a project to examine whether patterns can be identified between specified types of warranty claims and the QC data collected during the production process.

Companies need to analyse data carefully to establish the correct rules and distributions for their models.

In addition output from experimentation needs to be understood as effectively as possible.

Witness Miner helps to extract key information in the support of both of these tasks.

For Witness customers this offers a new way to analyse simulation input and output information.

Data can be analysed more effectively to provide more accurate rules to simulation models and the output from a range of experiments can be analysed where needed to determine the causes of success.

This includes the identification of patterns in data which suggest different structures of input data for Witness models, and an analysis of Witness Optimizer output data identifying causal relationships between input parameters and good simulation results.