163 research outputs found
Fight silent horror unit test methods by consulting a TestWizard
No full textTests, when not correctly implemented, can pass on incorrect system implementations rather than fail. In this case, they are named silent horrors or false-negative tests. They make releasing low-quality (buggy) versions of the software system more probable. Furthermore, faithfully implementing test specifications is crucial when they play the role of documentation, like when documenting components or services or driving legacy systems' re-engineering. This paper presents TestWizard, a novel approach and tool for automatically assessing individual tests' quality from the point of view of their coherence to specifications. TestWizard automatically assesses the quality of each individual test case w.r.t. its specification, providing detailed reports on why a single test is a false negative, hence helping testers fix them. Thus, TestWizard can help to automate the test code review process, which is still mainly manual today. The analysis of 1012 test implementations, developed by 123 students in three experiments, shows that TestWizard is (1) by far more accurate than code review performed by multiple students, (2) slightly better than code review performed by three senior experts, and (3) always able to detect a significant percentage of false-negative test methods (up to 21.22%)
Il punto di vista dell’ortopedico e del neurochirurgo di fronte alla frattura vertebrale nella fase d’urgenza. Problematiche a confronto”
No full textAbstract compilation of object-oriented languages into coinductive CLP(X): can type inference meet verification?
No full textThis paper further investigates the potential and
practical applicability of abstract compilation in two
different directions. First, we formally define an
abstract compilation scheme for precise prediction of
uncaught exceptions for a simple Java-like language;
besides the usual user declared checked exceptions, the
analysis covers the runtime ClassCastException. Second,
we present a general implementation schema for abstract
compilation based on coinductive CLP with variance
annotation of user-defined predicates, and propose an
implementation based on a Prolog prototype
meta-interpreter, parametric in the solver for the
subtyping constraints
Familial adult myoclonic epilepsy: A new expansion repeats disorder
No full textFamilial adult myoclonic epilepsy (FAME), also described with different acronyms (ADCME, BAFME, FEME, FCTE and others), is a high-penetrant autosomal dominant condition featuring cortical hand tremors, myoclonic jerks, and occasional/rare convulsive seizures. Prevalence is unknown since this condition is often under-recognized, but it is estimated to be less than 1/35,000. The disease usually starts in the second decade of life and has been genetically associated with at least 4 different loci (8q24, 2p11.1-q12.2, 5p15.31-p15 and 3q26.32−3q28). Recently, the expansion of non coding TTTTA and TTTCA repeats has been identified as the causative mutation in Japanese families linked to the 8q24. The diagnosis is supported by clinical features and electrophysiological investigations as jerk-locked back averaging, C-reflex, and somatosensory-evoked potential. Photic stimulation, emotional stress, and sleep deprivation may trigger both tonic-clonic and myoclonic seizures. FAME has a slow but progressive clinical course occurring with intellectual disability and worsening of both tremor and myoclonus although with a less severe decline compared to other progressive myoclonic epilepsies. Valproate, levetiracetam, and benzodiazepines are considered the first-line treatments
Optimizing and Evaluating Pre- Trained Large Language Models for Alzheimer's Disease Detection
No full textThis research focuses on developing improved diagnostic tools for Alzheimer's Disease (AD), a condition impacting approximately 50 million individuals globally. In the paper, we achieve automatic AD detection by leveraging pre-trained Large Language Models (LLMs) for linguistic analysis applied to the ADReSS/ADReSSo (Alzheimer's Dementia Recognition through Spontaneous Speech/only) Challenges datasets, following speech-to-text conversion. While the recent advancements in LLMs offer a robust foundation for their application in healthcare, fine-tuning these models for specific tasks, like AD detection, requires optimization to balance performance and computational efficiency. Also in response to data privacy concerns in healthcare, we implement our methodology on consumer-level GPU cards, which offer a practical solution for local data processing. Our approach uses fine-tuning techniques such as Low Ranking Adaptation and Parameter-Efficient Fine-Tuning to enhance the capabilities of Large Language Models within the limits of consumer-grade hardware. Additionally, we incorporate quantization to reduce computational demands while preserving model accuracy. Conducted on setups with RTX 4090 and dual RTX 3090 GPUs, our experiments demonstrate promising results that align with or surpass existing benchmarks in dementia recognition
- …
