Unsupervised machine learning is the machine learning task of inferring a function to describe a hidden structure from unlabeled data. We could state that unsupervised learning is a feature of natural neural networks that is required to be replicated in a machine toward the achievement of autonomous unsupervised learning. In order to achieve the outputs of unsupervised learning, the platform designed by Responsible Machines highlights several key features that assist a robot in attaining its self-aware state.
Data Organization: In order to realize unsupervised intelligence, it is important to organize data that is conducive to instant pattern recognition. This may be compared to the Adaptive Resonance Theory (ART), developed by Stephen Grossberg and Gail Carpenter. The basic ART system typically consists of a comparison field, a recognition field that is comprised of neurons, a vigilance parameter (threshold of recognition), and a reset module. This platform organizes data in a manner that simplifies identification and comparison tasks.
Pattern Recognition: The recorded dataset is stored in a fashion that creates a unique data set pattern. As the phrase goes “Cells that fire together, wire together”; data patterns are formed employing spatial and temporal markers. These data patterns are used to recognize and compare, and based on previous learning, occurrence or supervision, a specific weight is allocated to the relationship between these parameters.
String Creation (Tag Assembly): These data patterns comprise a linear array of values, which form strings. A sample string, which appears similar to the example the below, might contain identifiers for every micro-data set within the string.
Whenever a data pattern is fired, the string checks every micro-identifier for comparison. Parameters that agree with the match result in classification, and when combined with the parameters that disagree with the match, enable the creation of a unique label.
Based on repetitive or supervised occurrences, a linear weight is assigned to the relationship between data parameters. Utilizing these weights, the best match whose set of weights (weight vector) most closely matches the input vector is selected as the response.
This platform works very similarly to the ART mentioned above, in terms of how the brain processes information.