Browsing by Subject "Artificial neural network"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    An Integrated Nanocomposite Proximity Sensor: Machine Learning-Based Optimization, Simulation, and Experiment
    (MDPI, 2022-04-08) Moheimani, Reza; Gonzalez, Marcial; Dalir, Hamid; Mechanical and Energy Engineering, School of Engineering and Technology
    This paper utilizes multi-objective optimization for efficient fabrication of a novel Carbon Nanotube (CNT) based nanocomposite proximity sensor. A previously developed model is utilized to generate a large data set required for optimization which included dimensions of the film sensor, applied excitation frequency, medium permittivity, and resistivity of sensor dielectric, to maximize sensor sensitivity and minimize the cost of the material used. To decrease the runtime of the original model, an artificial neural network (ANN) is implemented by generating a one-thousand samples data set to create and train a black-box model. This model is used as the fitness function of a genetic algorithm (GA) model for dual-objective optimization. We also represented the 2D Pareto Frontier of optimum solutions and scatters of distribution. A parametric study is also performed to discern the effects of the various device parameters. The results provide a wide range of geometrical data leading to the maximum sensitivity at the minimum cost of conductive nanoparticles. The innovative contribution of this research is the combination of GA and ANN, which results in a fast and accurate optimization scheme.
  • Thumbnail Image
    Item
    Modeling dose-dependent temperature responses to methamphetamine
    (Springer Nature, 2012-07-16) Molkov, Yaroslav; Zaretsky, Dmitry; Zaretskaia, Maria; Rusyniak, Dan; Mathematical Sciences, School of Science
  • Thumbnail Image
    Item
    Predicting Bacterial Vaginosis Development using Artificial Neural Networks
    (medRxiv, 2025-05-05) Elnaggar, Jacob H.; Lammons, John W.; Ardizzone, Caleb M.; Aaron, Kristal J.; Jacobs, Clayton; Graves, Keonte J.; George, Sheridan D.; Luo, Meng; Tamhane, Ashutosh; Łaniewski, Paweł; Quayle, Alison J.; Herbst-Kralovetz, Melissa M.; Cerca, Nuno; Muzny, Christina A.; Taylor, Christopher M.; Microbiology and Immunology, School of Medicine
    Bacterial vaginosis (BV) is a dysbiosis of the vaginal microbiome, characterized by the depletion of protective Lactobacillus spp. and overgrowth of anaerobes. Artificial neural network (ANN) modeling of vaginal microbial communities offers an opportunity for early detection of incident BV (iBV). 16S rRNA gene sequencing and quantitative PCR was performed on longitudinal vaginal specimens collected from participants within 14 days of iBV or from healthy participants to calculate the inferred absolute abundance (IAA) of vaginal bacterial taxa. ANNs were trained using the IAA of vaginal taxa from 420 vaginal specimens to classify individual vaginal specimens as either pre-iBV (collected before iBV onset) or Healthy. Feature importance was assessed to understand how specific vaginal micro-organisms contributed to model predictions. ANN modeling accurately classified >97% of specimens as either pre-iBV or Healthy (sensitivity >96%, specificity >98%) using IAA of 20 vaginal taxa. Model prediction accuracy was maintained when training models using only a few key vaginal taxa. Models trained using only the top five most important features achieved an accuracy of >97%, sensitivity >92%, and specificity >99%. Model predictive accuracy was further improved by training models on specimens from white and black participants separately; using only three feature models achieved an accuracy >96%, sensitivity >91%, and specificity >91%. Feature analysis found that Lactobacillus species L. gasseri and L. jensenii differed in how they contributed to model predictions in models trained with data stratified by race. A total of 420 vaginal specimens were analyzed, providing a robust dataset for model training and validation.
  • Thumbnail Image
    Item
    Predicting future cancer burden in the United States by artificial neural networks
    (Taylor & Francis, 2021) Piva, Francesco; Tartari, Francesca; Giulietti, Matteo; Aiello, Marco Maria; Cheng, Liang; Lopez-Beltran, Antonio; Mazzucchelli, Roberta; Cimadamore, Alessia; Cerqueti, Roy; Battelli, Nicola; Montironi, Rodolfo; Santoni, Matteo; Pathology and Laboratory Medicine, School of Medicine
    Aims: To capture the complex relationships between risk factors and cancer incidences in the US and predict future cancer burden. Materials & methods: Two artificial neural network (ANN) algorithms were adopted: a multilayer feed-forward network (MLFFNN) and a nonlinear autoregressive network with eXogenous inputs (NARX). Data on the incidence of the four most common tumors (breast, colorectal, lung and prostate) from 1992 to 2016 (available from National Cancer Institute online datasets) were used for training and validation, and data until 2050 were predicted. Results: The rapid decreasing trend of prostate cancer incidence started in 2010 will continue until 2018-2019; it will then slow down and reach a plateau after 2050, with several differences among ethnicities. The incidence of breast cancer will reach a plateau in 2030, whereas colorectal cancer incidence will reach a minimum value of 35 per 100,000 in 2030. As for lung cancer, the incidence will decrease from 50 per 100,000 (2017) to 31 per 100,000 in 2030 and 26 per 100,000 in 2050. Conclusion: This up-to-date prediction of cancer burden in the US could be a crucial resource for planning and evaluation of cancer-control programs.