Abstract Motivation: Understanding the substrate specificity of human immunodeficiency virus (HIV)-1 protease is important when designing effective HIV-1 inhibitors. Furthermore, characterizing and predicting cleavage profile essential to generate test hypotheses how affects proteins host. Currently available tools for by can be improved. Results: The linear support vector machine with orthogonal encoding shown best predictor cleavage. It considerably better than current publicly services....

Abstract Summary: Several papers have been published where nonlinear machine learning algorithms, e.g. artificial neural networks, support vector machines and decision trees, used to model the specificity of HIV-1 protease extract rules. We show that dataset in these studies is linearly separable it a misuse classifiers apply them this problem. The best solution on achieved using linear classifier like simple perceptron or machine, straightforward rules from models. identify key residues...

Using a neural-network classifier we are able to separate gluon from quark jets originating Monte Carlo--generated ${\mathit{e}}^{+}$${\mathit{e}}^{\mathrm{\ensuremath{-}}}$ events with 85%--90% accuracy.

Abstract Summary: A set of new algorithms and software tools for automatic protein identification using peptide mass fingerprinting is presented. The automatic, fast modular to suit different laboratory needs, it can be operated either via a Java user interface or called from within scripts. modules do peak extraction, filtering database matching, communicate XML. Individual therefore easily replaced with other if desired, all intermediate results are available the user. designed operate...

The standard machine learning assumption that training and test data are drawn from the same probability distribution does not hold in many real-world applications due to inability reproduce testing conditions at time. Existing unsupervised domain adaption (UDA) methods address this problem by a domain-invariant feature space performs well on available source domain(s) (labeled data) specific target (unlabeled data). In contrast, instead of simply adapting domains, paper aims for an approach...

This paper investigates the novel application of Large Language Models (LLMs) with vision capabilities to analyze satellite imagery for village-level poverty prediction. Although LLMs were originally designed natural language understanding, their adaptability multimodal tasks, including geospatial analysis, has opened new frontiers in data-driven research. By leveraging advancements vision-enabled LLMs, we assess ability provide interpretable, scalable, and reliable insights into human from...

Rapidly developing viral resistance to licensed human immunodeficiency virus type 1 (HIV-1) protease inhibitors is an increasing problem in the treatment of HIV-infected individuals and AIDS patients. A rational design more effective discovery potential biological substrates for HIV-1 require accurate models cleavage specificity. In this study, several popular bioinformatic machine learning methods, including support vector machines artificial neural networks, were used analyze specificity...

Managing the maintenance of a commercial vehicle fleet is an attractive application domain ubiquitous knowledge discovery. Cost effective methods for predictive are progres- sively demanded in automotive industry. The traditional diagnostic paradigm that requires human experts to define models not scalable today's vehicles with hundreds computing units and thousands control sensor signals streaming through on-board controller area network. A more autonomous approach must be developed. In...

Vehicle utilization analysis is an essential tool for manufacturers to understand customer needs, improve equipment uptime, and collect information future vehicle service development. Typically today, this behavioral modeling done on high-resolution time-resolved data with features such as GPS position fuel consumption. However, costly transfer sensitive from a privacy perspective. Therefore, typically only collected when the pays extra services relying that data. This motivated us develop...

An approach for intelligent monitoring of mobile cyberphysical systems is described, based on consensus among distributed self-organised agents. Its usefulness experimentally demonstrated over a long-time case study in an example domain: fleet city buses. The proposed solution combines several techniques, allowing life-long learning under computational and communication constraints. presented work step towards autonomous knowledge discovery domain where data volumes are increasing, the...

A robust air/fuel ratio “soft sensor” is presented based on non-linear signal processing of the ion current using neural networks. Care taken to make system insensitive amplitude variations, due e.g. fuel additives, by suitable preprocessing signal.

The Langevin updating rule, in which noise is added to the weights during learning, presented and shown improve learning on problems with initially ill-conditioned Hessians. This particularly important for multilayer perceptrons many hidden layers, that often have In addition, Manhattan a similar effect.

Proteases of human pathogens are becoming increasingly important drug targets, hence it is necessary to understand their substrate specificity and interpret this knowledge in practically useful ways. New methods being developed that produce large amounts cleavage information for individual proteases some have been applied extract rules from data. However, the hitherto proposed extracting neither easy nor very accurate. To be useful, should accurate, compact, expressed an easily...

A telematic based system for enabling automatic fault detection of a population vehicles is proposed. To avoid sending huge amounts data over the telematics gateway, idea to use low-dimensional representations sensor values in sub-systems vehicle. These are then compared between similar systems fleet. If representation vehicle found deviate from group fleet, labeled diagnostics that subsystem. The demonstrated on engine coolant and it shown how this self-organizing approach can detect...

Two spark advance control systems are outlined; both based on feedback from nonlinear neural network soft sensors and ion current detection. One uses an estimate the location of pressure peak other center combustion. Both quantities estimated signal using networks. The estimates correct within roughly two crank angle degrees when evaluated a cycle to basis, one degree averaged over consecutive cycles.The detection system is demonstrated SAAB 9000 car, equipped with 2.3 liter low-pressure...

In order to maximize protein identification by peptide mass fingerprinting noise peaks must be removed from spectra and recalibration is often required. The preprocessing of the before database searching essential but time-consuming. Nevertheless, optimal search parameters vary over a batch samples. For high-throughput identification, these factors should set automatically, with no or little human intervention. present work automated filtering using statistical filter described. combined...