Internet of Things applications analyze our past habits through sensor measures to anticipate future trends. To yield accurate predictions, intelligent systems not only rely on single numerical values, but also on structured models aggregated from different sensors. Computation theory, based on the discretization of observable data into timed events, can easily lead to millions of values. Time series and similar database structures can efficiently index the mere data, but quickly reach computation and storage limits when it comes to structuring and processing IoT data. We propose a concept of continuous models that can handle high-volatile IoT data by defining a new type of meta attribute, which represents the continuous nature of IoT data. On top of traditional discrete object-oriented modeling APIs, we enable models to represent very large sequences of sensor values by using mathematical polynomials. We show on various IoT datasets that this significantly improves storage and reasoning efficiency.

Background/Objectives: Head and neck squamous cell carcinoma (HNSCC) diagnosis and treatment rely heavily on computed tomography (CT) imaging to evaluate tumor characteristics and lymph node (LN) involvement, crucial for staging, prognosis, and therapy planning. Conventional LN evaluation methods based on morphological criteria such as size, shape, and anatomical location often lack sensitivity for early metastasis detection. This study leverages radiomics to extract quantitative features from CT images, addressing the limitations of subjective assessment and aiming to enhance LN classification accuracy while potentially reducing reliance on invasive histopathology. Methods: We analyzed 234 LNs from 27 HNSCC patients, deriving 120 features per node, resulting in over 25,000 data points classified into reactive, pathologic, and pathologic with extracapsular spread classes. Considering the challenges of high dimensionality and limited dataset size, more than 44,000 experiments systematically optimized machine learning models, feature selection methods, and hyperparameters, including ensemble approaches to strengthen classification robustness. A Pareto front strategy was employed to balance diagnostic accuracy with significant feature reduction. Results: The optimal model, validated via 5-fold cross-validation, achieved a balanced accuracy of 0.90, an area under the curve (AUC) of 0.93, and an F1-score of 0.88 using only five radiomics features. Conclusions: This interpretable approach aligns well with clinical radiology practice, demonstrating strong potential for clinical application in noninvasive LN classification in HNSCC.

The advent of Internet of Things will revolutionise the sharing mobility by enabling high connectivity between passengers and means of transport. This generates enormous quantity of data which can reveal valuable knowledge and help understand complex travel behaviour. At the same time, it challenges analytics platforms to discover knowledge from data in motion (i.e., the analytics occur in real time as the event happens), extract travel habits, and provide reliable and faster sharing mobility services in dynamic contexts. In this paper, a scalable method for dynamic profiling is introduced, which allows the extraction of users’ travel behaviour and valuable knowledge about visited locations, using only geolocation data collected from mobile devices. The methodology makes use of a compact representation of time-evolving graphs that can be used to analyse complex data in motion. In particular, we demonstrate that using a combination of state-of-the-art technologies from data science domain coupled with methodologies from the transportation domain, it is possible to implement, with the minimum of resources, the next generation of autonomous sharing mobility services (i.e., long-term and on-demand parking sharing and combinations of car sharing and ride sharing) and extract from raw data, without any user input and in near real time, valuable knowledge (i.e., location labelling and activity classification).

Many real world computer vision applications are required to run on hardware with limited computing power, often referred to as "edge devices". The state of the art in computer vision continues towards ever bigger and deeper neural networks with equally rising computational requirements. Model compression methods promise to substantially reduce the computation time and memory demands with little to no impact on the model robustness. However, evaluation of the compression is mostly based on theoretic speedups in terms of required floating-point operations. This work offers a tool to profile the actual speedup offered by several compression algorithms. Our results show a significant discrepancy between the theoretical and actual speedup on various hardware setups. Furthermore, we show the potential of model compressions and highlight the importance of selecting the right compression algorithm for a target task and hardware. The code to reproduce our experiments is available at https://hub.datathings.com/papers/2022-coins.

Data-driven model interpretability is a requirement to gain the acceptance of process engineers to rely on the prediction of a data-driven model to regulate industrial processes in the ironmaking industry. In the research presented in this paper, we focus on the development of an interpretable multivariate time series forecasting deep learning architecture for the temperature of the hot metal produced by a blast furnace. A Long Short-Term Memory (LSTM) based architecture enhanced with attention mechanism and guided backpropagation is proposed to accommodate the prediction with a local temporal interpretability for each input. Results are showing high potential for this architecture applied to blast furnace data and providing interpretability correctly reflecting the true complex variables relations dictated by the inherent blast furnace process, and with reduced prediction error compared to a recurrent-based deep learning architecture.