Over-performance technology based on hypercubic analysis

Data-mining refers to the set of calculation techniques used to extract information from databases. This terminology brings together a wide range of concepts, from correlations to artificial intelligence by way of clustering.

These techniques, which have been used regularly for many years, share a base in statistical theory. Statistical techniques are used primarily for sampling and require only a small amount of calculation power, which encouraged their development in the twentieth century.

However, the statistical approach describes only the overall influence of data under study. Based on ranking, elimination or regression techniques, statistical data mining is very long and gives insufficient answers in most cases. In addition, actual physical phenomena seldom correspond to the necessary prerequisites for high-quality statistical analysis (normality, consistency, linearity). Most users have little confidence in these technologies since they are so difficult to master.

Compared to all other approaches, Braincube represents a major breakthrough because it includes a hypercubic analysis engine with analytical power that goes well beyond statistics.

Hypercubic analysis uses neither statistical methods nor derivations of these methods. The Braincube search engine analyzes phenomena under study by pinpointing local “over-performance.” In contrast to a global approach, hypercubic analysis is particularly suited to nonlinear systems involving threshold effects, like those encountered in a real environment.

Although hypercubic analysis requires significant computing power, Braincube’s innovative architecture allows deploying this technology in any organization, with no need for major investment.

Using hypercubic analysis for industrial processes enables obtaining results that would be impossible with statistical analysis theories.