Standard feature engineering involves manually designing measurable descriptors based on some expert knowledge in the domain of application, followed by the selection of the best performing set of designed features for the subsequent optimisation of an inference model. Several studies have shown that this whole manual process can be efficiently replaced by deep learning approaches which are characterised by the integration of feature engineering, feature selection and inference model optimisation into a single learning process. In the following work, deep learning architectures are designed for the assessment of measurable physiological channels in order to perform an accurate classification of different levels of artificially induced nociceptive pain. In contrast to previous works, which rely on carefully designed sets of hand-crafted features, the current work aims at building competitive pain intensity inference models through autonomous feature learning, based on deep neural networks. The assessment of the designed deep learning architectures is based on the BioVid Heat Pain Database (Part A) and experimental validation demonstrates that the proposed uni-modal architecture for the electrodermal activity (EDA) and the deep fusion approaches significantly outperform previous methods reported in the literature, with respective average performances of $84.57\%$ and $84.40\%$ for the binary classification experiment consisting of the discrimination between the baseline and the pain tolerance level ( $T_0$ vs. $T_4$ ) in a Leave-One-Subject-Out (LOSO) cross-validation evaluation setting. Moreover, the experimental results clearly show the relevance of the proposed approaches, which also offer more flexibility in the case of transfer learning due to the modular nature of deep neural networks.

Ordinal classification (OC) is an important niche of supervised pattern recognition, in which the classes constitute an ordinal structure. In general, the ordinal structure can be identified, either according to the natural occurrence of the current task (e.g. healthy - mild condition - moderate condition - severe condition), or by extracting expert knowledge. However, we assume that many multi-class classification tasks might have a hidden ordinal structure, which, once identified, can facilitate and hence leverage the classification process. Therefore, we propose a working definition for OC tasks, which is based on the decision boundaries of standard binary Support Vector Machines. Moreover, resulting from our proposed definition, we introduce a simple algorithm for the detection of ordinal structures. Our proposed definition is easy to interpret and reflects an intuitive understanding of ordinal structures. Another main advantage is that our proposed definition is easy to apply. Therefore, there is no more dependence on expert knowledge for the identification of (non-intuitive) ordinal class structures. In the current study, we include ten benchmark data sets from the field of OC to experimentally evaluate and hence to confirm the validity of our proposed definition. Additionally, we analyse our proposed definition based on a small set of traditionally non-ordinal multi-class classification tasks. Furthermore, we provide an analysis of the computational cost of our proposed detection algorithm, and discuss the limitations of our proposed working definition.

Standard feature engineering involves manually designing and assessing measurable descriptors based on some expert knowledge in the domain of application, followed by the selection of the best performing set of designed features in order to optimize an inference model. Several studies have shown that this whole manual process can be efficiently replaced by deep learning approaches which are characterized by the integration of feature engineering, feature selection and inference model optimization into a single learning process. Such techniques have proven to be very successful in the domain of image processing and have been able to attain state-of-the-art performances while significantly outperforming traditional approaches based on hand-crafted features. In the following work, we explore deep learning approaches for the analysis of physiological signals. More precisely, deep learning architectures are designed for the assessment of measurable physiological channels in order to perform an accurate classification of different levels of artificially induced nociceptive pain. Most of the previous works related to pain intensity classification based on physiological signals rely on a carefully designed set of hand-crafted features in order to achieve a relatively good classification performance. Therefore, the current work aims at building competitive pain intensity classification models without the need of domain specific expert knowledge for the generation of relevant features. The assessment of the designed deep learning architectures is based on the BioVid Heat Pain Database (Part A) and experimental validation demonstrates that the proposed uni-modal architecture for the electrodermal activity (EDA) and the deep fusion approaches significantly outperform previous classification methods reported in the literature, with respective average performances of 85.03% and 83.76% for the binary classification experiment consisting of the discrimination between the baseline level and the pain tolerance level ( T ₀ vs . T ₄ ) in a Leave - One - Subject - Out (LOSO) cross-validation evaluation setting.

In general, classification tasks can differ significantly in their task complexity. For instance, image-based differentiation between vehicles and pedestrians is most likely expected to be less complex than CT-scan-based differentiation between several lung diseases. Intuitively, based on a human point of view, one can identify some classification tasks as more complex than other classification tasks. Moreover, based on expert knowledge and/or task-specific meta information, one could attempt to estimate the complexity ranks of specific classification tasks. In this work, based on the publicly available BioVid Heat Pain Database (BVDB), we experimentally confirm the intuitive assumption that the task of automated pain intensity recognition (PIR) is very challenging. Inspired by the field of chaos theory, we show that the BVDB-specific PIR task can not only be seen as highly complex, but is even identified as a classification task of chaotic nature. To this end, we apply Hao’s working definition for chaotic systems and provide an experiment-based chaos check method. To validate our approach, as a non-complex counterpart, we include a task of handwritten numerals distinction. Our study provides two main contributions, i.e.: i) an enhanced understanding for the still present and – more importantly – substantial gap between the ground truth and the predictions reported by different research groups in combination with automated PIR tasks; and ii) an approach for a numerical complexity check based on chaos theory. Different research directions are discussed for future work. Note that improving PIR accuracy performance is not part of the study objective.