Intelligensys automatic data mining, machine learning for product formulation

Historically, developing rules has been the province of the human expert. And, typically, learning the rules in any application domain requires a long apprenticeship. However, when you are faced with a new knowledge domain, today's fast-paced world means that you don't have the luxury of spending time to develop rules in the conventional manner.

Automatic rule generation - sometimes called data mining - is a way of extracting rules directly from data and presenting them in an easily understood format. Some of the techniques that are used include:

neurofuzzy logic as used in FormRules
CART - Classification And Regression Trees. This includes algorithms like ID3, C4.5 and C5

More details about these are given in the on-line PowerPoint presentation, given at the January 2000 PROFITS meeting.

Intelligensys Solutions
To see some examples of data mining in action, have a look at our Application Examples. You'll see that sensible rules have been extracted automatically from data sets taken from the open literature.

These examples were carried out using FormRules, our data mining package based on neurofuzzy logic. FormRules has been designed to be used by formulators themselves, without requiring them to be 'artificial intelligence experts'.

Don't want to do it yourself, but want the results? We can carry out the work for you, on a completely confidential basis. Just provide us with your data, and our contract research team can extract the rules that will soon have you formulating faster and better. To discuss this in more detail and find out how it could work for you, just contact us. And of course we can use other techniques - like decision trees - to look at your data too, so that you get maximum value.