Embedded Files
Synthetic Cognition has been develop to be able to solve any problem. It can be applied for simple prediction or regression tasks, for knowledge discovery, or even for complex reinforcement learning agent problems. Moreover, eventually we will be able to apply Synthetic Cognition to solve tasks that require an Artificial General Intelligence, although we are still not there.
With this broad spectrum of possibilities, we have been able to apply Synthetic Cognition in fields as diverse as business analytics, image recognition, agent control, and health data exploration.
Knowledge Discovery
We have developed tools to explore Synthetic Cognition's knowledge representation. They allow to discover new knowledge hidden in the data.
We have tools to perform Pattern Mining, Feature Selection and Dimensionality Reduction using the knowledge representation of a trained Synthetic Cognition.
Prediction and Regression
We have adapted Synthetic Cognition to be able to perform prediction and regression, with additional tools to explain the decisions made.
Synthetic Cognition can perform Prediction and Regression at the level of current state-of-the-art Machine Learning algorithms.
Reinforcement Learning Agents
We have prepared Synthetic Cognition to learn from mistake, to be able to model and reproduce behavior.
Synthetic Cognition acts as an Agent with its own policy, needing lower training costs to achieve similar performance of current state-of-the-art Reinforcement Learning algorithms.
Additionally, these agents can be purely reactive agents or episodic agents, thanks to Synthetic Cognition capabilities.
Artificial General Intelligence Agents
The final goal of Synthetic Cognition is to be able to develop agents that have planning and reasoning capabilities similar to humans. However, this is pretty much work in progress right now, thus we still can not offer this kind of agents.
Our Successful Applications:
We have successfully applied our technology in the following use-cases:
HIV therapeutic vaccine trial transcriptomic data analysis for Aelix Therapeutics.
Gut Microbiota Link to Sexual Preference and HIV Infection for IrsiCaixa.
Response to Immune Checkpoint Inhibitors in Advanced Solid Tumors for Vall d’Hebron Institute of Oncology (VHIO).
We have also achieve state-of-the-art results in the following benchmarks:
Predictive Models for Breast Cancer Classification based on multi-omic data.
Report abuse