bayesian-neural-network

Bayesian Neural Network

blog

website related stuff

blogtheme

cnn-tutorial

A Convolution Neural Network (CNN) Tutorial

deeplearning

deeplearning LSTM adjusted

digital-twin-approach-for-damage-tolerant-mission-planning-under-uncertainty

The digital twin paradigm that integrates the information obtained from sensor data, physics models, as well as operational and inspection/maintenance/repair history of a system (or a component) of interest, can potentially be used to optimize operational parameters of the system in order to achieve a desired performance or reliability goal. In this article, we develop a methodology for intelligent mission planning using the digital twin approach, with the objective of performing the required work while meeting the damage tolerance requirement. The proposed approach has three components: damage diagnosis, damage prognosis, and mission optimization. All three components are affected by uncertainty regarding system properties, operational parameters, loading and environment, as well as uncertainties in sensor data and prediction models. Therefore the proposed methodology includes the quantification of the uncertainty in diagnosis, prognosis, and optimization, considering both aleatory and epistemic uncertainty sources. We discuss an illustrative fatigue crack growth experiment to demonstrate the methodology for a simple mechanical component, and build a digital twin for the component. Using a laboratory experiment that utilizes the digital twin, we show how the trio of probabilistic diagnosis, prognosis, and mission planning can be used in conjunction with the digital twin of the component of interest to optimize the crack growth over single or multiple missions of fatigue loading, thus optimizing the interval between successive inspection, maintenance, and repair actions.

information-fusion-and-machine-learning-for-sensitivity-analysis-using-physics-knowledge-and-experim

Highlights • Physics-informed machine learning is investigated for global sensitivity analysis. • Physics and test data are fused to maximize the accuracy of sensitivity estimates. • Uncertainties in Gaussian process and deep neural network models are included. • Accuracy, uncertainty and computational effort of proposed approaches are compared.

kapusuzoglu

root repo

kapusuzoglu.github.io

my blog

machine-learning-with-python

Practice and tutorial-style notebooks covering wide variety of machine learning techniques

mfi

Joel Greenblatt - Magic Formula Investing

minillama

Replicated the architecture of Llama to create miniLlama for testing / demo purposes.

montecarlo-simluation-value-investing

multi-objective-optimization-under-uncertainty-of-part-quality-in-fused-filament-fabrication

Multi-objective optimization

numerai

Contributions

pca-tutorial

Principal Component Analysis (PCA)

personal-weight-predictions-using-machine-learning

What can machine learning tell you about your weight?

physics-informed-and-hybrid-machine-learning-in-additive-manufacturing

Physics-Informed and Hybrid Machine Learning in Additive Manufacturing: Application to Fused Filament Fabrication

piml_gsa

process-optimization-under-uncertainty-for-improving-the-bond-quality-of-polymer-filaments-in-fused-

This paper develops a computational framework to optimize the process parameters such that the bond quality between extruded polymer filaments is maximized in fused filament fabrication (FFF). A one-dimensional heat transfer analysis providing an estimate of the temperature profile of the filaments is coupled with a sintering neck growth model to assess the bond quality that occurs at the interfaces between adjacent filaments. Predicting the variability in the FFF process is essential for achieving proactive quality control of the manufactured part; however, the models used to predict the variability are affected by assumptions and approximations. This paper systematically quantifies the uncertainty in the bond quality model prediction due to various sources of uncertainty, both aleatory and epistemic, and includes the uncertainty in the process parameter optimization. Variance-based sensitivity analysis based on Sobol' indices is used to quantify the relative contributions of the different uncertainty sources to the uncertainty in the bond quality. A Gaussian process (GP) surrogate model is constructed to compute and include the model error within the optimization. Physical experiments are conducted to show that the proposed formulation for process parameter optimization under uncertainty results in high bond quality between adjoining filaments of the FFF product.

python-save-plots

Save plots as PDF/PGF (Progressive Graphics File) for direct upload to LaTeX.

riskfolio_lib

umami

Umami is a simple, fast, privacy-focused alternative to Google Analytics.

uncertainty-quantification

This document is prepared based on the lectures and notes for the graduate level course- ‘Uncertainty Quantification’ (CE 6310) taught by Dr. Sankaran Mahadevan at Vanderbilt University. Course materials of ‘Spring 2015’ and ‘Spring 2019’.