"... I THOUGHT OF CALLING IT ‘INFORMATION’, BUT THE WORD WAS OVERLY USED, SO I DECIDED TO CALL IT ‘UNCERTAINTY’ ..."

At AiGENT-TECH, we create Strong AI Agents (AiGENTs).
Using a novel blend of AI technologies - probabilistic models and inference engines driven by information entropy combined with Bayesian & Deep Reinforcement Learning - our AiGENTs learn faster, with smaller data set. Resulting superior AI Agents, shortest TTM and training efforts.

STRONG AI AGENTS

Cognitive digital workforce

The definition of (human) intelligence is fuzzy,

"A very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience ... "

, and all the more so when it comes to Artificial Intelligence (AI).

Strong AI, a long-term goal of research, is the attempt to give machines the complete set of human intelligence, including:

  • Reason, solve problems, and make judgments under uncertainty

  • Represent knowledge

  • Plan

  • Learn

  • Communicate in natural language

  • and integrate all the above skills towards a common goal.

Weak AI, in contrast, does not attempt to perform the above full range of human cognitive abilities. 

At AiGENT-TECH, we develop strong AI agents (AiGENTs) that perceive their environment and take actions which maximize their chances of success.

DEEP REINFORCEMENT LEARNING

Exploration exploitation dilemma optimization

Unlike supervised learning, reinforcement learning doesn't use input/output pairs.
Instead the focus is on on-line active learning, which involves finding a balance between exploration (of uncharted territory) and exploitation (of current knowledge).
Our multi-step horizon deep reinforcement learning algorithms maximize mission success rate, while minimizing mission cost function, resulting efficient & productive AiGENTs

RATIONAL COMPUTING

Machine learning approaches to human brain function

Combining:

  •  Statistical approaches to model human brain function (e.g. the cognitive abilities based on statistical principles).

  • Cloud based strong CPUs/GPUs, large memory pools.

  • Advanced machine learning algorithms.

our AiGENTs excel at any task.

Meeting human expert standards, AiGENT-TECH rational computing SW engine learns and adapts based on experience.

'ON LINE' ACTIVE LEARNING

Learning based on experience

Active learning is a special case in which a learning algorithm is able to interactively query the user (or any other information source) to obtain the desired output at a new scenario.
With this approach, there is a risk that the AiGENT (= the algorithm) be overwhelmed by uninformative examples and will miss his target.
AiGENT-TECH rational computing SW engine combines adaptive machine learning concepts with incremental learning policies to avoid overwhelmed situations (like the human brain does)

©2017 by AiGENT-Tech.com