Browse Articles
Discover research articles across all indexed journals
Hybrid clustering strategies for effective oversampling and undersampling in multiclass classification
Abstract Multiclass imbalance is a challenging problem in real-world datasets, where certain classes may have a low number of samples because they correspond to rare occurrences. To address the challenge of multiclass imbalance, this paper introduces a novel hybrid cluster-based oversampling and undersampling (HCBOU) technique. By clustering and separating classes into majority and minority categories, this algorithm retains the most information during undersampling while generating efficient data in the minority class. The classification is carried out using one-vs-one and one-vs-all decomposition schemes. Extensive experimentation was carried out on 30 datasets to evaluate the proposed algorithm's performance. The results were subsequently compared with those of several state-of-the-art algorithms. Based on the results, the proposed algorithm outperforms the competing algorithms under different scenarios. Finally, The HCBOU algorithm demonstrated robust performance across varying class imbalance levels, highlighting its effectiveness in handling imbalanced datasets.
Penicillin allergy as an instrumental variable for estimating antibiotic effects on resistance
Dual-energy CT-derived virtual noncalcium imaging to assess bone marrow lesions in patients with knee osteoarthritis
Automatic speech recognition predicts contemporaneous earthquake fault displacement
Abstract Significant progress has been made in probing the state of an earthquake fault by applying machine learning to continuous seismic waveforms. The breakthroughs were originally obtained from laboratory shear experiments and numerical simulations of fault shear, then successfully extended to slow-slipping faults. Here we apply the Wav2Vec-2.0 self-supervised framework for automatic speech recognition to continuous seismic signals emanating from a sequence of moderate magnitude earthquakes during the 2018 caldera collapse at the Kīlauea volcano on the island of Hawai’i. We pre-train the Wav2Vec-2.0 model using caldera seismic waveforms and augment the model architecture to predict contemporaneous surface displacement during the caldera collapse sequence, a proxy for fault displacement. We find the model displacement predictions to be excellent. The model is adapted for near-future prediction information and found hints of prediction capability, but the results are not robust. The results demonstrate that earthquake faults emit seismic signatures in a similar manner to laboratory and numerical simulation faults, and artificial intelligence models developed for encoding audio of speech may have important applications in studying active fault zones.
The lean body mass to visceral fat mass ratio is negatively associated with cardiometabolic disorders: a cross-sectional study
Capacitance enhancement by ion-laminated borophene-like layered materials
A novel label-free method to determine equilibrium dissociation constants of antibodies binding to cell surface proteins
Author Correction: Multi-omic and single-cell profiling of chromothriptic medulloblastoma reveals genomic and transcriptomic consequences of genome instability
Determinants maintaining healthcare personnel’s motivation during COVID-19 pandemic in Uganda
The transcriptional response of cortical neurons to concussion reveals divergent fates after injury
Abstract Traumatic brain injury (TBI) is a risk factor for neurodegeneration, however little is known about how this kind of injury alters neuron subtypes. In this study, we follow neuronal populations over time after a single mild TBI (mTBI) to assess long ranging consequences of injury at the level of single, transcriptionally defined neuronal classes. We find that the stress-responsive Activating Transcription Factor 3 (ATF3) defines a population of cortical neurons after mTBI. Using an inducible reporter linked to ATF3, we genetically mark these damaged cells to track them over time. We find that a population in layer V undergoes cell death acutely after injury, while another in layer II/III survives long term and remains electrically active. To investigate the mechanism controlling layer V neuron death, we genetically silenced candidate stress response pathways. We found that the axon injury responsive dual leucine zipper kinase (DLK) is required for the layer V neuron death. This work provides a rationale for targeting the DLK signaling pathway as a therapeutic intervention for traumatic brain injury. Beyond this, our approach to track neurons after a mild, subclinical injury can inform our understanding of neuronal susceptibility to repeated impacts.
Pulsed laser deposition assisted epitaxial growth of cesium telluride photocathodes for high brightness electron sources
Investigating the Antiscale Magnetic Treatment Controversy: Insights from the Model Calcium Carbonate Scalant
Abstract The antiscale magnetic treatment (ASMT) claims to utilize magnetic field to combat scaling. However, its underlying mechanism, effectiveness, and reliability remain controversial. To address these contentious aspects, we analyze the influence of a magnetic field on the different stages of typical scale formation, using $${\text{CaCO}}_{3}$$ as a model scale. For simplification, we consider the working fluid, such as in domestic and industrial settings, as a homogeneous mixture of a supersaturated, multi-ionic solution and a suspension of neutral multiphase contaminants, a fraction of which is magnetic. We argue that the combined effects of pH variation and catalytic role of magnetic contaminants are crucial factors affecting the properties of the resultant scale. Based on these considerations, we clarify the controversy by showing that each side holds a valid piece of the overall picture of the ASMT process. Indeed, the two viewpoints on magnetic field’s influence on scaling can be explained along the following scenarios: (i) Within a non-contaminated, supersaturated solution, there is no significant field influence because, under typical laboratory conditions, the Lorentz force does not practically affect the scaling process. (ii) Within a high-pH, magnetically-contaminated, supersaturated solution, the field does have an influence: Here, gradient-force-driven agglomerated particulates can act as templates for heterogeneous nucleation and growth.
Gut–brain axis and neuropsychiatric health: recent advances
Alu-Sc-mediated exonization generated a mitochondrial LKB1 gene variant found only in higher order primates
Sensitivity of coronary hemodynamics to vascular structure variations in health and disease
Nanosilicon application changes the morphological attributes and essential oil compositions of hemp (Cannabis sativa L.) under water deficit stress
A bioinformatic approach to characterize the vitellogenin receptor and the low density lipoprotein receptor superfamily in the newt Cynops orientalis
Abstract The Low Density Lipoprotein receptors (LDLRs) gene family includes 15 receptors: very low-density lipoprotein receptor (VLDLR), LDLR, Sorting-related receptor with A-type repeats (SORLA), and 12 LDL receptor-related proteins (LRPs): LRP1, LRP1B, LRP2, LRP3, LRP4, LRP5, LRP6, LRP8, LRP10, LRP11, LRP12, LRP13. Most of these are involved in the transduction of key signals during embryonic development and in the regulation of cholesterol homeostasis. In oviparous animals, the VLDL receptor is also known as VTGR since it facilitates the uptake of vitellogenin in ovary. In tetrapods, information concerning genes encoding these proteins is limited to a few taxa. Here, we report the characterization of VTGR in the amphibian Cynops orientalis. The secondary structure analyses and the expression profiles obtained from hepatic and gonadal tissues of C. orientalis supported the role of VTGR as vitellogenin oocyte membrane receptor in this species. Moreover, to get a holistic view of the evolutionary history of this gene superfamily, we extended our investigation to all 15 genes belonging to the LDLR superfamily analyzing through a phylogenetic analysis a total of 161 sequences belonging to 11 genera of vertebrates. The position of LRP8 in the tree and its expression findings in C. orientalis ovary allowed us to suggest that other proteins of the LDLR superfamily could act as receptors during vitellogenesis.