A life-long learning vector quantization approach for interactive learning of multiple categories

Abstract

We present a new method capable of learning multiple categories in an interactive and lifelong learning fashion to approach the "stability-plasticity dilemma". The problem of incremental learning of multiple categories is still largely unsolved. This is especially true for the domain of cognitive robotics, requiring real-time and interactive learning. To achieve the life-long learning ability for a cognitive system, we propose a new learning vector quantization approach combined with a category-specific feature selection method to allow several metrical "views" on the representation space of each individual vector quantization node. These category-specific features are incrementally collected during the learning process, so that a balance between the correction of wrong representations and the stability of acquired knowledge is achieved. We demonstrate our approach for a difficult visual categorization task, where the learning is applied for several complex-shaped objects rotated in depth.

Combining offline and online classifiers for life-long learning

“…We use the learning technique proposed in [23] that has gained much attention recently in the context of big data and interpretable models due to its flexibility and intuitive classification scheme, see e. g. [33], [34], [35], [36], [37], [38], [39], [40], [41], [42]. Essentially, it offers an efficient way for prototypebased data classification.…”

“…Due to its representation of data in terms of prototypes, efficient incremental learning schemes have been proposed [37], [38], [39], [42], [46], [47], [49]. These schemes rely on the fact that prototypes can be used as a summary statistics to represent the already seen data, such that incremental online adaptation is easily possible based on the learned prototypes and new data.…”

Metric learning for sequences in relational LVQ

“…However, metric learning can be integrated efficiently into the classification scheme, and results from learning theory can be derived by referring to the resulting function class. This has been demonstrated in the context of learning vector quantization (LVQ), where metric learning opened the way towards efficient state-of-the-art results in various areas, including biomedical data analysis, robotic vision, and spectral analysis [4,19,1]. Because of the intuitive definition of models in terms of prototypical representatives, prototype-based methods like LVQ enjoy a wide popularity in application domains, particularly if human inspection and interaction are necessary, or life-long model adaptation is considered [29,20,18].…”

Autonomous Learning of Representations

“…in biomedical data analysis, on the one side, and life-long learning, on the other side [42,10,17,13,33,61]. This is caused by the fact that LVQ inherently relies on a representation of the model in terms of typical prototypes such that an intutitive model inspection as well as a compact data description is delivered by the technique; when referring to a probabilistic formalization of LVQ techniques, this corresponds to data being represented by a mixture of Gaussians -however, unlike classical Gaussian mixture models, the parameters are optimized according to the conditional likelihood of the output label rather than the data itself [44].…”