Unsupervised Machine Learning

Store buildings as tourist attractions: Mining retail meaning of store building pictures through a machine learning approach

“…This cluster analysis allows isolation of different groups (clusters) within a sample by examining the common features (Perez and Nadal, 2005), and is largely used in the analyses of pictures (Nikjoo and Bankhshi, 2019). The process of clustering can be considered as an unsupervised data-analysis technique employed to elicit hidden patterns in datasets (Hodeghatta and Nayak, 2017), where the researchers do not provide the algorithm with input of how to identify the cluster (no training), but rather, the algorithm discovers patterns in the data (image in this case) automatically (Balducci and Marinova, 2018). Such machine learning techniques are especially effective when working with nonnumeric data (highly unstructured data) (Balducci and Marinova, 2018).…”

Clustering Techniques for Traffic Classification: A Comprehensive Review

“…Machine Learning (ML) approaches are advanced in terms of accuracy, performance, and complexity compared to traditional approaches. ML enables a system to train itself with existing database, from which the system later infers appropriate decisions regarding the traffic classification [31]. The approach of learning falls under three groups specifically supervised, unsupervised and semi-supervised.…”

“…The unlabeled data, however, forms the majority of the dataset fed as input to the system. Clustering techniques fall under unsupervised and semi supervised learning and mostly has to deal with association of some characteristic features [31]. There are different methods of clustering, namely classic K-means, hierarchical, density-based, grid-based, probabilistic model-based, and hybrid models [14], [17], [18], [32].…”