Outlier detection means to characterize the distribution of inliers exactly enough such that outliers stand out. A natural and interpretable model is an enclosing ball that includes non-outliers but excludes outliers. However, a single enclosing ball is insufficient for data with multiple, distinct clusters or modes. Therefore, we propose a model with K enclosing balls, each of which can cover a separate cluster of data. We investigate two optimization schemes, one based on stochastic gradient descent and one based on alternating optimization and quadratic programming. On synthetic and real-world data we compare our proposed K minimum enclosing ball (K-MEB) approach to baselines from the literature.

In this work we propose an white-box workflow for regression related tasks based on near-infrared data in the context of machine learning. The workflow consists of data pre-processing and dimension reduction through (inverse) Fourier-transformation and low-pass filtering of the spectra. Subsequently, a machine learning model shall be applied to predict food contents of the spectra and conclude the workflow. To yield recommendations, we test various standard models, as well as the iterative method of partial-least-squares and the recently proposed Regression (Sensitive) Neural Gas. We shall not only investigate performance aspects, but also discuss theoretical concepts, i.e. interpretability options offered by each model. We show that our pre-processing and reduction approach is able to achieve good results even for signal-to-noise ratio dependent models and that the Regression (Sensitive) Neural Gas offers rich options to gain insights into the data and the model results.

We consider the so-called symmetric structural auto-encoder (SyS-AE) for image reconstruction preserving the perceptual properties to be kept in the decoded images for later visual inspection by experts. The images are obtained from micro-gravity experiments on the International Space Station (ISS). This application requires an encoder of low computational complexity and fast execution time due to the limited hardware and energy resources. The proposed SyS-AE uses a non-linear transfer function on vectorized images followed by a linear down-projection as encoder trained by stochastic gradient descent learning using the structural similarity loss. The decoder can be explicitly calculated as a kind of an inverse of the encoder. We demonstrate the ability of the system for the given problem as well as for illustrating MNIST and Fashion-MNIST data.

. This paper is an effort to parameterize Information Bottle-neck Theory to become a supervised classifier. We introduce a parametrization by means of Learning Vector Quantization. With this new approach, one can find suitable components that are necessary for an accurate, yet efficient, classification. A balance between compression and representation is made by means of a specially designed objective function.

We present a method, which allows to train a Generalized Matrix Learning Vector Quantization (GMLVQ) model for classification using data from several, maybe non-calibrated, sources without explicit transfer learning. This is achieved by using a siamese-like GMLVQ-architecture, which comprises different sets of prototypes for the target classification and for the separation learning of the sources. In this architecture, a linear map is trained by means of GMLVQ for source distinction in the mapping space in parallel to the classification task learning. The respective null-space projection provides a common data representation of the different source data for an all-together classification learning.

One-class-classification remains an important problem in machine learning, which is related to data representation and outlier detection, but different from them in several aspects. In the present contribution we propose an one-class-classifier based on a prototype vector quantization model. We modeled a corresponding cost function to account for aspects of representation learning and to appropriately evaluate the one-class classifier. The prototype -based model ensures a local representation of the target class. After this introduction, we obtain an interpretable one-class classifier model. We demonstrate the capabilities of the approach by applying the classifier to illustrative toy data examples as well as on real data in a medical context.

This paper contains a feasibility study of deep neural networks for the classification of Euro banknotes with respect to requirements of central banks on the ATM and high speed sorting industry. Instead of concentrating on the accuracy for a large number of classes as in the famous ImageNet Challenge we focus thus on conditions with few classes and the requirement of rejection of images belonging clearly to neither of the trained classes (i.e. classification in a so-called 0-class). These special requirements are part of frameworks defined by central banks as the European Central Bank and are met by current ATMs and high speed sorting machines. We also consider training and classification time on state of the art GPU hardware. The study concentrates on the banknote recognition whereas banknote class dependent authenticity and fitness checks are a topic of its own which is not considered in this work.