UbiQTree: Uncertainty quantification in XAI with tree ensembles
UbiQTree quantifies uncertainty in SHAP explanations for tree ensemble models, decomposing it into aleatoric (irreducible) and epistemic (reducible) componen...
Publication Categories
- Bioinformatics 9
- Machine-Learning 9
- Data-Visualization 4
- Genomics 3
- AI-Ethics 3
- User-Interfaces 2
- Time-series 2
- Molecular-Modeling 2
- Peptide-Classification 2
- AMR 2
- Explainable-AI 2
- Storage-Technologies 1
- Model-Evaluation 1
- Control-Systems 1
- Large-Language-Models 1
Bioinformatics
Genomic data representations for horizontal gene transfer detection
RCKmer (k=7) + SVM representation achieves state-of-the-art HGT detection (F1=0.959, MCC=0.908) across 44 genomic encodings and 4 datasets, outperforming tra...
Protocol for implementing the nested model for AI design and validation in compliance with AI regulations
Protocol for implementing the nested model to develop/validate AI systems compliant with regulations, spanning domain, data, model, and prediction layers wit...
Current state and future prospects of Horizontal Gene Transfer detection
Review reveals surging AI adoption in Horizontal Gene Transfer (HGT) detection—a key driver of prokaryotic evolution and antimicrobial resistance (AMR)—with ...
Interpretable Molecular Encodings and Representations for Machine Learning Tasks
iCAN delivers interpretable carbon-neighborhood encodings for organic molecules, outperforming prior methods in peptide/protein classification across large b...
Interactive polar diagrams for model comparison
polar-diagrams is an interactive library implementing Mutual Information and Taylor Diagrams to visualize multi-model performance across biology, medicine, c...
A parametric approach for molecular encodings using multilevel atomic neighborhoods applied to peptide classification
We propose cmangoes, a parametric, carbon-chain-based encoding method for organic molecules that generates binary/numerical arrays (exportable as images) fro...
MOVIS: A multi-omics software solution for multi-modal time-series clustering, embedding, and visualizing tasks
MOVIS is a modular, web-based tool for exploring time-series multi-omics data, designed to make complex temporal, multi-modal datasets accessible to research...
The Visual Story of Data Storage: From Storage Properties to User Interfaces
Data creation is rapidly outpacing traditional storage, driving the need for new media such as DNA storage. This work surveys user requirements across key st...
Machine-Learning
UbiQTree: Uncertainty quantification in XAI with tree ensembles
UbiQTree quantifies uncertainty in SHAP explanations for tree ensemble models, decomposing it into aleatoric (irreducible) and epistemic (reducible) componen...
Genomic data representations for horizontal gene transfer detection
RCKmer (k=7) + SVM representation achieves state-of-the-art HGT detection (F1=0.959, MCC=0.908) across 44 genomic encodings and 4 datasets, outperforming tra...
Persona Adaptable Strategies Make Large Language Models Tractable
Ten persona-adaptable strategies enable large language models to self-explain their functionality to everyday users, tailoring complexity/style to user arche...
Protocol for implementing the nested model for AI design and validation in compliance with AI regulations
Protocol for implementing the nested model to develop/validate AI systems compliant with regulations, spanning domain, data, model, and prediction layers wit...
Current state and future prospects of Horizontal Gene Transfer detection
Review reveals surging AI adoption in Horizontal Gene Transfer (HGT) detection—a key driver of prokaryotic evolution and antimicrobial resistance (AMR)—with ...
Kernel-based Learning for Safe Control of Discrete-Time Unknown Systems under Incomplete Observations
Kernel ridge regression enables safe tracking control for high-order dynamical systems with unknown dynamics and partial state observations. Our observer-int...
Interpretable Molecular Encodings and Representations for Machine Learning Tasks
iCAN delivers interpretable carbon-neighborhood encodings for organic molecules, outperforming prior methods in peptide/protein classification across large b...
Interactive polar diagrams for model comparison
polar-diagrams is an interactive library implementing Mutual Information and Taylor Diagrams to visualize multi-model performance across biology, medicine, c...
A parametric approach for molecular encodings using multilevel atomic neighborhoods applied to peptide classification
We propose cmangoes, a parametric, carbon-chain-based encoding method for organic molecules that generates binary/numerical arrays (exportable as images) fro...
Data-Visualization
Persona Adaptable Strategies Make Large Language Models Tractable
Ten persona-adaptable strategies enable large language models to self-explain their functionality to everyday users, tailoring complexity/style to user arche...
Interactive polar diagrams for model comparison
polar-diagrams is an interactive library implementing Mutual Information and Taylor Diagrams to visualize multi-model performance across biology, medicine, c...
MOVIS: A multi-omics software solution for multi-modal time-series clustering, embedding, and visualizing tasks
MOVIS is a modular, web-based tool for exploring time-series multi-omics data, designed to make complex temporal, multi-modal datasets accessible to research...
The Visual Story of Data Storage: From Storage Properties to User Interfaces
Data creation is rapidly outpacing traditional storage, driving the need for new media such as DNA storage. This work surveys user requirements across key st...
Genomics
Genomic data representations for horizontal gene transfer detection
RCKmer (k=7) + SVM representation achieves state-of-the-art HGT detection (F1=0.959, MCC=0.908) across 44 genomic encodings and 4 datasets, outperforming tra...
Current state and future prospects of Horizontal Gene Transfer detection
Review reveals surging AI adoption in Horizontal Gene Transfer (HGT) detection—a key driver of prokaryotic evolution and antimicrobial resistance (AMR)—with ...
MOVIS: A multi-omics software solution for multi-modal time-series clustering, embedding, and visualizing tasks
MOVIS is a modular, web-based tool for exploring time-series multi-omics data, designed to make complex temporal, multi-modal datasets accessible to research...
AI-Ethics
UbiQTree: Uncertainty quantification in XAI with tree ensembles
UbiQTree quantifies uncertainty in SHAP explanations for tree ensemble models, decomposing it into aleatoric (irreducible) and epistemic (reducible) componen...
Persona Adaptable Strategies Make Large Language Models Tractable
Ten persona-adaptable strategies enable large language models to self-explain their functionality to everyday users, tailoring complexity/style to user arche...
Protocol for implementing the nested model for AI design and validation in compliance with AI regulations
Protocol for implementing the nested model to develop/validate AI systems compliant with regulations, spanning domain, data, model, and prediction layers wit...
User-Interfaces
MOVIS: A multi-omics software solution for multi-modal time-series clustering, embedding, and visualizing tasks
MOVIS is a modular, web-based tool for exploring time-series multi-omics data, designed to make complex temporal, multi-modal datasets accessible to research...
The Visual Story of Data Storage: From Storage Properties to User Interfaces
Data creation is rapidly outpacing traditional storage, driving the need for new media such as DNA storage. This work surveys user requirements across key st...
Time-series
Interactive polar diagrams for model comparison
polar-diagrams is an interactive library implementing Mutual Information and Taylor Diagrams to visualize multi-model performance across biology, medicine, c...
MOVIS: A multi-omics software solution for multi-modal time-series clustering, embedding, and visualizing tasks
MOVIS is a modular, web-based tool for exploring time-series multi-omics data, designed to make complex temporal, multi-modal datasets accessible to research...
Molecular-Modeling
Interpretable Molecular Encodings and Representations for Machine Learning Tasks
iCAN delivers interpretable carbon-neighborhood encodings for organic molecules, outperforming prior methods in peptide/protein classification across large b...
A parametric approach for molecular encodings using multilevel atomic neighborhoods applied to peptide classification
We propose cmangoes, a parametric, carbon-chain-based encoding method for organic molecules that generates binary/numerical arrays (exportable as images) fro...
Peptide-Classification
Interpretable Molecular Encodings and Representations for Machine Learning Tasks
iCAN delivers interpretable carbon-neighborhood encodings for organic molecules, outperforming prior methods in peptide/protein classification across large b...
A parametric approach for molecular encodings using multilevel atomic neighborhoods applied to peptide classification
We propose cmangoes, a parametric, carbon-chain-based encoding method for organic molecules that generates binary/numerical arrays (exportable as images) fro...
AMR
Genomic data representations for horizontal gene transfer detection
RCKmer (k=7) + SVM representation achieves state-of-the-art HGT detection (F1=0.959, MCC=0.908) across 44 genomic encodings and 4 datasets, outperforming tra...
Current state and future prospects of Horizontal Gene Transfer detection
Review reveals surging AI adoption in Horizontal Gene Transfer (HGT) detection—a key driver of prokaryotic evolution and antimicrobial resistance (AMR)—with ...
Explainable-AI
UbiQTree: Uncertainty quantification in XAI with tree ensembles
UbiQTree quantifies uncertainty in SHAP explanations for tree ensemble models, decomposing it into aleatoric (irreducible) and epistemic (reducible) componen...
Persona Adaptable Strategies Make Large Language Models Tractable
Ten persona-adaptable strategies enable large language models to self-explain their functionality to everyday users, tailoring complexity/style to user arche...
Storage-Technologies
The Visual Story of Data Storage: From Storage Properties to User Interfaces
Data creation is rapidly outpacing traditional storage, driving the need for new media such as DNA storage. This work surveys user requirements across key st...
Model-Evaluation
Interactive polar diagrams for model comparison
polar-diagrams is an interactive library implementing Mutual Information and Taylor Diagrams to visualize multi-model performance across biology, medicine, c...
Control-Systems
Kernel-based Learning for Safe Control of Discrete-Time Unknown Systems under Incomplete Observations
Kernel ridge regression enables safe tracking control for high-order dynamical systems with unknown dynamics and partial state observations. Our observer-int...
Large-Language-Models
Persona Adaptable Strategies Make Large Language Models Tractable
Ten persona-adaptable strategies enable large language models to self-explain their functionality to everyday users, tailoring complexity/style to user arche...