Background: There are no exact data about the prevalence of familial chylomicronemia syndrome (FCS) in Central Europe. We aimed to identify FCS patients using either the FCS score proposed by Moulin et al. or with data mining, and assessed the diagnostic applicability of the FCS score. Methods: Analyzing medical records of 1,342,124 patients, the FCS score of each patient was calculated. Based on the data of previously diagnosed FCS patients, we trained machine learning models to identify other features that may improve FCS score calculation. Results: We identified 26 patients with an FCS score of ≥10. From the trained models, boosting tree models and support vector machines performed the best for patient recognition with overall AUC above 0.95, while artificial neural networks accomplished above 0.8, indicating less efficacy. We identified laboratory features that can be considered as additions to the FCS score calculation. Conclusions: The estimated prevalence of FCS was 19.4 per million in our region, which exceeds the prevalence data of other European countries. Analysis of larger regional and country-wide data might increase the number of FCS cases. Although FCS score is an excellent tool in identifying potential FCS patients, consideration of some other features may improve its accuracy.

To better understand the potential of soils, understanding how soil properties vary over time and in-field is essential to optimize the cultivation and site-specific technologies in crop production. This article aimed at determining the within-field mapping of soil chemical and physical properties, vegetation index, and yield of maize in 2002, 2006, 2010, 2013, and 2017, respectively. The objectives of this five-year field study were: (i) to assess the spatial and temporal variability of attributes related to the maize yield; and (ii) to analyse the temporal stability of management zones. The experiment was carried out in a 15.3 ha research field in Hungary. The soil measurements included sand, silt, clay content (%), pH, phosphorous (P₂O₅), potassium (K₂O), and zinc (Zn) in the topsoil (30 cm). The apparent soil electrical conductivity was measured in two layers (0–30 cm and 30–90 cm, mS/m) in 2010, in 2013, and in 2017. The soil properties and maize yields were evaluated in 62 management zones, covering the whole research area. The properties were characterized as the spatial-temporal variability of these parameters and crop yields. Classic statistics and geostatistics were used to analyze the results. The maize yields were significantly positively correlated (r = 0.62–0.73) with the apparent electrical conductivity (Veris_N3, Veris_N4) in 2013 and 2017, and with clay content (r = 0.56–0.81) in 2002, 2013, and 2017.

Recommendation services bear great importance in e-commerce, shopping, tourism, and social media, as they aid the user in navigating through the items that are most relevant to their needs. In order to build recommender systems, organizations log the item consumption in their user sessions by using different sensors. For instance, Web sites use Web data loggers, museums and shopping centers rely on user in-door positioning systems to register user movement, and Location-Based Social Networks use Global Positioning System for out-door user tracking. Most organizations do not have a detailed history of previous activities or purchases by the user. Hence, in most cases recommenders propose items that are similar to the most recent ones viewed in the current user session. The corresponding task is called session based, and when only the last item is considered, it is referred to as item-to-item recommendation. A natural way of building next-item recommendations relies on item-to-item similarities and item-to-item transitions in the form of “people who viewed this, also viewed” lists. Such methods, however, depend on local information for the given item pairs, which can result in unstable results for items with short transaction history, especially in connection with the cold-start items that recently appeared and had no time yet to accumulate a sufficient number of transactions. In this paper, we give new algorithms by defining a global probabilistic similarity model of all the items based on Random Fields. We give a generative model for the item interactions based on arbitrary distance measures over the items, including explicit, implicit ratings and external metadata to estimate and predict item-to-item transition probabilities. We exploit our new model in two different item similarity algorithms, as well as a feature representation in a recurrent neural network based recommender. Our experiments on various publicly available data sets show that our new model outperforms simple similarity baseline methods and combines well with recent item-to-item and deep learning recommenders under several different performance metrics.

Recent articles indicate that deep neural networks are efficient models for various learning problems. However, they are often highly sensitive to various changes that cannot be detected by an independent observer. As our understanding of deep neural networks with traditional generalisation bounds still remains incomplete, there are several measures which capture the behaviour of the model in case of small changes at a specific state. In this paper we consider Gaussian perturbations in the tangent space and suggest tangent sensitivity in order to characterise the stability of gradient updates. We focus on a particular kind of stability with respect to changes in parameters that are induced by individual examples without known labels. We derive several easily computable bounds and empirical measures for feed-forward fully connected ReLU (Rectified Linear Unit) networks and connect tangent sensitivity to the distribution of the activation regions in the input space realised by the network.

By Bálint Daróczy, Tamás Kiss & 4 more. In this article we give a comprehensive overview of features devised for web spam detection and investigate how much various classes, some requiring very high computational effort,...

Introduction: The integration of AI into healthcare is widely anticipated to revolutionize medical diagnostics, enabling earlier, more accurate disease detection and personalized care.Methods: In this study, we developed and validated an AI-assisted diagnostic support tool using only routinely ordered and broadly available blood tests to predict the presence of major chronic and acute diseases as well as rare disorders.Results: Our model was tested on both retrospective and prospective datasets comprising over one million patients. We evaluated the diagnostic performance by (1) implementing ensemble learning (mean ROC-AUC.9293 and mean DOR 63.96); (2) assessing the model's sensitivity via risk scores to simulate its screening effectiveness; (3) analyzing the potential for early disease detection (30–270 days before clinical diagnosis) through creating historical patient timelines and (4) conducting validation on real-world clinical data in collaboration with Synlab Hungary, to assess the tool's performance in clinical setting.Discussion: Uniquely, our model not only considers stable blood values but also tracks changes from baseline across 15 years of patient history. Our AI-driven automated diagnostic tool can significantly enhance clinical practice by recognizing patterns in common and rare diseases, including malignancies. The models' ability to detect diseases 1–9 months earlier than traditional clinical diagnosis could contribute to reduced healthcare costs and improved patient outcomes. The automated evaluation also reduces evaluation time of healthcare providers, which accelerates diagnostic processes. By utilizing only routine blood tests and ensemble methods, the tool demonstrates high efficacy across independent laboratories and hospitals, making it an exceptionally valuable screening resource for primary care physicians.