Sinus Iridum - Knowledge from Confusion

How does the Uncertainty Engine Work?

The UE sees patterns in data in the same way that a human mind recognises patterns, but the UE has the capacity for phenomenal size and complexity (trials included Genome data modelling for example).

The Uncertainty Engine works with patterns, not rules. No two sets of circumstances are ever the same, so a rules-based program always runs the risk of missing something, whilst at the same time using up resources on irrelevancies. At the outset of a project, the Engine will seek to discover patterns in data. From there it can identify direct matches with patterns elsewhere, and also correlations where there might be inexact matches, and because capacity is not an issue for the Engine, the more comprehensive the data, the more accurate the output. At this stage, it will identify any patterns of interest, which SI will bring to a client's attention, and a decision can be made as to relevance, relevance which, because the Engine retains knowledge, can then be used to finesse the process as it continues.

Working in the binary, the engine is data agnostic and holds everything as being uncertain until it becomes certain. Every new piece of data updates all previous results and nothing is excluded (unlike traditional systems that need to reduce volume and complexity to facilitate processing), as something that may be irrelevant now, could become relevant in the future.

In describing the UE capability, direct life experiences are often recalled by clients when they suddenly see the logic behind the UE approach and appreciate the phenomenal capability of the engine; we present just two instances for your interest:

The father of a five year old said that he visited a supermarket to buy salad items. His daughter picked up a courgette and announced that she had found the cucumbers.

The second story recounted a family walk in Yorkshire hills, when a young child having a ride on his father's shoulders saw some deer in the distance. "Look daddy, Rabbits; big ones".

You can understand the logic in both stories as they clearly demonstrate powerful pattern matching capability of developing young minds seeking to build experience; uncertain until becoming certain.

An exact parallel to the engine.

If you are only interested in finding cucumbers or rabbits, then it is relatively easy to directly match both, but to identify all of the permutations that 'might' be rabbits or cucumbers is way beyond the capabilities of traditional direct pattern match technologies even those employing mathematically induced fuzzy edges to those matches. Hence the traditional supplier question "What would you like us to look for?"

Direct matching is worth considering further. Imagine the programming involved in just describing a rabbit, or even a cucumber - so many permutations and that description would probably be a representation of just the visual. The children will have used multiple senses (dimensions) to evaluate their target; as does the UE - 5 in the case of the children and 4000 in the case of the UE; with each UE dimension capable of containing 4000 entities – as was required by the genome project.