Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 55 for “"Feature Importance"”.
-
Interpretable Deep Learning: Beyond Feature-Importance with Concept-based Explanations
… DNN model behaviour to improve interpretability. Feature importance explanations are the most popular interpretability approaches. They show the importance of each input feature (e.g., pixel, patch, word vector) to the model’s prediction. However, we hypothesise that feature importance …
-
Unconfined compressive strength prediction using drilling parameters and analyzing feature importance through principal components analysis
… Component Analysis (PCA) to indicate the importance of each parameter by quantifying their variance contribution. Random Forest machine learning algorithm is utilized to build a regression model to estimate UCS. The regression model developed uses Depth, Rotation per Minute (RPM), …
-
ENHANCING pm2.5 AIR POLLUTION ANALYSIS IN BEIJING: A TRANSITION FROM NON-PARAMETRIC REGRESSION TO ADVANCED MACHINE LEARNING METHODOLOGY
… frequency, to assess PM2.5 levels as well as feature importance, may result in overfitting, lack of interpretability and computational inefficiency. Our research aims to address these limitations by making use of the RF+ framework, due its robustness in high-dimensional data and its ability to …
-
A machine learning approach to predict emotional arousal and aggressive driving from EDA and heart rate signals
… activity (EDA) and Heart Rate signals, 22 features were extracted, including tonic and phasic components, difference-based features, and heart rate statistics. Feature selection was performed using Random Forest feature importance and L1-regularized Logistic Regression feature importance …
-
Informing decision-making in single-objective, mixed-variable design problems
… technique determining the directional importance of mixed variables in a design space is benchmarked against state-of-the-art variable importance methods (also known as feature importance or interpretability) from machine learning. The importance evaluations and runtimes are compared …
-
Explainable AI: A Unified Approach Based on Cooperative Game Theory
… of machine learning (ML) models, yet existing feature-based explanation methods remain fragmented across local and global approaches. This thesis presents a unified framework based on cooperative game theory to systematically characterize and distinguish feature-based explanations. By …
-
Interpretability of Neural Networks Latent Representations
… We focus on 3 types of interpretability methods: feature importance, example-based and concept-based explanations. Most feature importance methods require a label to select which component of the neural network output to interpret. When interpreting neural networks representations, components …
-
Interpretable Deep Learning for Time Series
… of interpretability methods for detecting feature importance over time using both precision and recall. We find that network architectures and saliency methods fail to reliably and accurately identify feature importance over time. For RNNs, saliency vanishes over time, biasing detection of …
-
Cheap talk and costly consequences
… in trade. We implemented and applied a new feature selection algorithm called Boruta to build a compact set of predictor variables for this task. After finding a consistent set of features we used a Random Forest Classifier to predict bilateral changes in trade between 1998-2014. To better …
-
Optimal feature selection and machine learning for high-level audio classification : a random forests approach
… is predominately dependent on the selected features. This thesis presents a detailed study to identify the suitable classification features and associate a suitable machine learning technique for the intended classification task. In particular, a systematic feature selection procedure is …
-
SYNTHETIC OVERSAMPLING IMPROVES SPECTRAL DETECTION OF AFLATOXIN IN SINGLE MAIZE KERNELS
… the overall distribution of AF contamination. Feature importance distributions overlapped among models at 329-345 nm, 380-385.5 nm, 415-425 nm, 639-668 nm, and 1,013.5-1,060 nm. These spectral ranges could be applicable to the development of limited wavelength grain sorting devices for …
-
A Deep Learning Framework for Investigating Spatio-temporal Evolution of Land Use and Land Cover Patterns
… as the base learners to handle spatio-temporal features and incorporates an attention mechanism to indicate feature importance. The proposed prediction method is also tested for simulating likely scenarios with different urban expansion rates. The research contributes to the advancement of …
-
Seemingly Unrelated Research in Economics
… These drivers were confirmed by quantifying feature importance using extreme gradient boosting. Estimating a mediation model of the change in affect on satisfaction with life showed these results continued to hold when controlling for personality traits and income. Our analyses show the …
-
A Data-Driven Approach for Predicting and Understanding Braking Conditions of Aircraft Landings
… action to all cases with reduced friction. Feature importance is computed using SHAP values, showing that relative humidity, temperature, precipitation, and aircraft type are the features that guide model predictions the most. The model can be used to create decision aids for aircraft …
-
Encoder-Agnostic Learned Temporal Matching for Video Classification
… fails to account for a variety of time-related features, such as variable video durations, chronological order of events, and temporal variance in feature significance. While methods for temporal modeling do exist, they often require significant architectural changes and expensive retraining, …
-
AI-POWERED DIGITAL TWIN MODELING USING LARGE HIGHWAY INFRASTRUCTURE DATASETS
… landslide risk assessment, earth surface feature detection by instance segmentation and dynamic FEM simulations for structural behavior under environmental and operational loads.In this study, machine learning models were trained on extensive historical highway datasets to predict …
-
Integrated Machine Learning and Bioinformatics Approaches for Prediction of Cancer-Driving Gene Mutations
… of cancer drivers and were the most informative features in the driver mutation classification. Through extensive comparative analysis with structure-functional experiments and multicenter mutational calling data from PanCancer Atlas studies, we have demonstrated the robustness of our models and …
Page 1 of 3