What can I use this for?

This interface helps you work with situations in which there are several possible explanations, states, locations, or scenarios — and new evidence becomes available.

The basic question is simple:

Given what we believed before, how should new evidence change the relative weight of the possibilities?

Spherion uses the BAYES operation to perform that update while preserving the exact computational rules of the Kernel.

A simple example

Suppose there are four possible explanations for an observation.

You already have information about those possibilities. Then a new observation arrives and is more compatible with some of them than with others.

For example, the new evidence might have relative likelihoods such as:

0.8 · 0.6 · 0.2 · 0.1

This does not automatically mean that the first possibility is the answer.

The new evidence is combined with the previous state. Spherion then produces the resulting updated state.

Where can this be useful?

Medical and scientific diagnosis
Several hypotheses may explain an observation. A new test or measurement provides additional evidence, changing their relative plausibility. In medical contexts, such calculations would support analysis rather than replace clinical judgment.
Weather and climate analysis
Different weather scenarios can be assigned prior weights. New measurements from radar, satellites, or weather stations can change the relative support for each scenario.
Robotics and localization
A robot may have several possible locations. A new sensor reading can make some locations more compatible with the observation than others.
Astronomy
Several hypotheses may describe an observed object. New measurements — such as spectral, brightness, or motion data — can change the relative weight of those hypotheses.
Geology and geophysical exploration
Different models of underground structures can be compared with new seismic, gravitational, or other measurements.
Ecology
Alternative hypotheses about the presence, distribution, or behavior of a species can be updated when new observations or sensor data become available.
Industrial monitoring
Several possible conditions may explain the behavior of a machine or process. New sensor readings can change the relative support for normal operation, wear, sensor problems, or other conditions.
Economic and scenario analysis
Several competing scenarios can be maintained simultaneously. New observations can change their relative weights without treating any single scenario as certain.

What Spherion is doing

The interface is not asking you to choose the answer in advance.

You provide:

  • the existing state;
  • the new evidence.

Spherion calculates how that evidence changes the state according to the current Kernel rules.

Before anything is executed, the interface shows you what will be used and asks for explicit authorization.

About the numbers you enter

You should enter ordinary mathematical values, such as:

0.8

You do not need to specify internal formats such as f64.

The interface is responsible for converting the value into the exact computational representation required by SWI while preserving the value you entered.

Before execution, you will be able to confirm the values that will actually be used.

This interface will evolve

This is an experimental interface to the current capabilities of the Spherion Kernel.

New examples, explanations, controls, visualizations, and other forms of interaction may be added in future versions.