Adaptive Mixture Discriminant Analysis for Supervised Learning with Unobserved Classes

Abstract

International audienceIn supervised learning, an important issue usually not taken into account by classical methods is the possibility of having in the test set individuals belonging to a class which has not been observed during the learning phase. Classical supervised algorithms will automatically label such observations as belonging to one of the known classes in the training set and will not be able to detect new classes. This work introduces a model-based discriminant analysis method, called adaptive mixture discriminant analysis (AMDA), which is able to detect several unobserved groups of points and to adapt the learned classifier to the new situation. Two EM-based procedures are proposed for parameter estimation and model selection criteria are used for selecting the actual number of classes. Experiments on artificial and real data demonstrate the ability of the proposed method to deal with complex and real word problems. The proposed approach is also applied to the detection of unobserved communities in social network analysis

A two-stage Bayesian semiparametric model for novelty detection with robust prior information

“…By letting these two parameters go off to infinity, we can also recover the degenerate case P T r j = δˆ j where the Dirac's delta denotes a point mass centered inˆ j . That is, the prior beliefs extracted from the training set can be flexibly updated by gradually transitioning from transductive to inductive inference by increasing λ T r and ν T r (Bouveyron 2014). Similarly, we set H ≡ N I W (m 0 , λ 0 , ν 0 , S 0 ) , where the hyperparameters are chosen to induce a flat prior for the novel components.…”

“…In the training set, we identify as outliers units with implausible labels and/or values. Cappozzo et al (2020) extend the work of Bouveyron (2014) addressing this problem by using a robust estimator that relies on impartial trimming (Gordaliza 1991). In short, the most unlikely data points under the currently estimated model are discarded.…”

The Application of Gaussian Mixture Models for the Identification of At-Risk Learners in Massive Open Online Courses

“…Accordingly, an EDDA model is designed and evaluated for the prediction of student risk within course environments, as considered within the hard categories of "at risk", or "not at risk". The supervised classification algorithms did not take into consideration the impact of unlabeled data on one of class [22]. EDDA capable autonomously discover unobserved latent engagement and assign these unlabeled data to one of the classes.…”

“…EDDA capable autonomously discover unobserved latent engagement and assign these unlabeled data to one of the classes. The mixture model is powerful inference framework can approximate represented high dimensional data as a linear combination of multiple Gaussian components [22] [23].…”

Machine learning techniques for monitoring the sludge profile in a secondary settler tank

“…Applications of GMM in data monitoring can also be found in the literature, for example in sensor monitoring (Zhu et al 2014), fault detection and diagnosis (Jiang et al 2016;Yu 2012). Some other applications include data classification (Bouveyron 2014), image segmentation (Greggio et al 2011) and many others.…”