Unintentional falls can cause severe injuries and even death, especially if no immediate assistance is given. A fall detection system aims to detect a fall as soon as it occurs, therefore issuing an automatic assistance request. Wearable embedded sensors are emerging as the most viable solution for continuous monitoring since they are more effective, less intrusive and less expensive than other systems. Tailoring a deep learning method to the requirements of microcontrollers entails matching very stringent constraints in terms of both memory and computational power. In addition, datasets acquired with wearable devices are relatively scarce and not necessarily devised for supervised learning. In this work, we discuss the design of a software architecture based on recurrent neural networks which can be effective for fall detection while running entirely onboard a wearable device. The well-known and publicly-available SisFall dataset was adopted and extended with fine-grained temporal annotations to cope with the supervised training of recurrent neural networks. We then show that an appropriate process of architectural minimization together with accurate hyperparameters selection leads to a workable model which compares favorably with other detection techniques. The embedding of the resulting architecture has been validated using a state-of-art hardware device.

Understanding the pharmacokinetics, safety and efficacy of candidate drugs is crucial for their success. One key aspect is the characterization of absorption, distribution, metabolism, excretion and toxicity (ADMET) properties, which require early assessment in the drug discovery and development process. This study aims to present an innovative approach for predicting ADMET properties using attention-based graph neural networks (GNNs). The model utilizes a graph-based representation of molecules directly derived from Simplified Molecular Input Line Entry System (SMILE) notation. Information is processed sequentially, from substructures to the whole molecule, employing a bottom-up approach. The developed GNN is tested and compared with existing approaches using six benchmark datasets and by encompassing regression (lipophilicity and aqueous solubility) and classification (CYP2C9, CYP2C19, CYP2D6 and CYP3A4 inhibition) tasks. Results show the effectiveness of our model, which bypasses the computationally expensive retrieval and selection of molecular descriptors. This approach provides a valuable tool for high-throughput screening, facilitating early assessment of ADMET properties and enhancing the likelihood of drug success in the development pipeline.

This study investigates the potential of ChatGPT 4 in the assessment of personality traits based on written texts. Using two publicly available datasets containing both written texts and self-assessments of the authors’ psychological traits based on the Big Five model, we aimed to evaluate the predictive performance of ChatGPT 4. For each sample text, we asked for numerical predictions on an eleven-point scale and compared them with the self-assessments. We also asked for ChatGPT 4 confidence scores on an eleven-point scale for each prediction. To keep the study within a manageable scope, a zero-prompt modality was chosen, although more sophisticated prompting strategies could potentially improve performance. The results show that ChatGPT 4 has moderate but significant abilities to automatically infer personality traits from written text. However, it also shows limitations in recognizing whether the input text is appropriate or representative enough to make accurate inferences, which could hinder practical applications. Furthermore, the results suggest that improved benchmarking methods could increase the efficiency and reliability of the evaluation process. These results pave the way for a more comprehensive evaluation of the capabilities of Large Language Models in assessing personality traits from written texts.