top of page

Decision Intelligence

1 December 2020

Elad Piran

The convergence of Data Science and the field of decision making led to the birth of a new approach to managing knowledge and information. Decision Intelligence's objective is to convert raw data into informed actions through an intelligent decision-making process.


About the approach

Decision Intelligence is a machine learning and AI approach. It claims that an advanced computing or AI system can acquire information items and perform decision-making process based on said system. This process would incorporate automatic learning processes and convert said decisions into informed actions. No human interference would be necessary, yet the process would resemble human decision-making processes.

The innovative aspect of this field is the AI's ability to acquire large amounts of information items and make decisions based on them. Furthermore, the AI's ability to learn and improve its decision-making process the more examples it receives.


How does it work?

When identifying a character of a cat in a picture, human beings will succeed regardless of its position. It can be standing and staring into the camera, lying down horizontally, or even viewed from behind. Since people are familiar with the concept of a cat, they can catalogue different pictures of a cat as a cat nonetheless.


A simple  computer, on the other hand, would identify the cat in the picture only if programmed in advance to identify a certain character of cat. It will therefore identify a character as a cat only if it highly matches the originally programmed picture. This means that if the picture it was programmed to identify as a cat consists of a cat photographed frontally, it will not identify a cat in any other position.


However, an advanced computing system complete with AI, once programed and performed an automatic learning process scanning additional examples of cat pictures, will identify a cat character regardless of its position. This includes angles not included in its initial programming.


How can this technology be applied?

The vast variety of uses and applications this feature of computerized data identification and learning is possible endless.

First and foremost, it can be used for basic technical functions, as applied by industries in several operating systems:

  • Face-identifying street cameras

  • Car-identifying parking lot cameras

  • Tracking devices for Quality Control purposes in factories

  • Face identification to be used in Smartphone applications (some of you probably unlock your phone by smiling to the frontal camera, don't you?)

  • Smart House applications

It is worth noting that this technology is not limited to the visual input, such as cameras and picture identification. While these were easy examples, this approach has reached any area or field in which AI can "learn", i.e. receive input and data and reach conclusions accordingly. There are currently AI systems that analyze written and digital texts, audio data, etc.


And the future? What about the future?

AI experts are dreaming aloud, describing an age in which computers will be able to make decisions and perform autonomously in nearly every field.

Maybe, one day, they will outdo us humans. 


bottom of page