Humans throughout history have been the entity that had the ability to reason abstractly about events as they unfold. No matter how instinctive animals can be, human intelligence has always had the super powers of reasoning. This interesting information came to us from MIT News in their article, “Toward a machine learning model that can reason about everyday actions.”
Computers, on the other hand, have a challenging time organizing the world into abstract categories. In recent years researchers have gotten closer by training machine learning models on words and images infused with structural information about the world and objects and animals within it.
As deep neural networks become expert at recognizing objects and actions in photos and video, researchers have set their sights on abstraction and training models to reason about what they see.
A deep learning model that can be trained to “think” more abstractly may be capable of learning with fewer data. Abstraction also paves the way toward higher-level, more human-like reasoning.
Data Harmony is a fully customizable suite of software products designed to maximize precise, efficient information management and retrieval. Our suite includes tools for taxonomy and thesauri construction, machine aided indexing, database management, information retrieval and explainable artificial intelligence.
Melody K. Smith
Sponsored by Data Harmony, harmonizing knowledge for a better search experience.
