,allowExpansion)
Quando la logica incontra l'apprendimento
Neuro symbolic artificial intelligence (Ne-Sy AI) is an emerging approach to artificial intelligence that combines the use of neural networks with symbolic AI: in other words, it exploits the powerful approximation and learning abilities of neural networks combined with symbolic reasoning, to manipulate abstract concepts, extrapolate information from limited amounts of data and generate explainable results.
The use of the neuro-symbolic approach is advantageous for several reasons: first of all, Ne-Sy AI systems are more efficient, reliable and robust, also because they require a smaller amount of training data than other models. In addition, the decision-making process has improved and the need for complex ML models is reduced, thanks to the integration of multiple logic and reasoning mechanisms.
)
Although this is a growing field, there are many possible uses of neuro Symbolic AI: from autonomous driving to recommender systems, from planning models to medical diagnoses. Among the most interesting cases, financial analyses constitute an important example, because neurosymbolic models are able to combine symbolic knowledge of the different regulations with machine learning models to improve market forecasting, detect fraud and manage financial risk. The benefit is the increased accuracy of forecasts and the reduction of risks associated with financial decisions.
Also from an industrial point of view, we are witnessing the birth of various neuro symbolic AI frameworks: one of these is Relational AI, which has the advantage of being integrated into Snowflake.