deep-ensemble

[TIP 2022] Towards Better Accuracy-efficiency Trade-offs: Divide and Co-training. Plus, an image classification toolbox includes ResNet, Wide-ResNet, ResNeXt, ResNeSt, ResNeXSt, SENet, Shake-Shake, DenseNet, PyramidNet, and EfficientNet.

Some example projects that was made using Tensorflow (mostly). This repository contains the projects that I've experimented-tried when I was new in Deep Learning.

Dataset and code for "Uncertainty-Informed Deep Transfer Learning of PFAS Toxicity"

Uncertainty Estimation Using Deep Neural Network and Gradient Boosting Methods

PyTorch implementation of Probabilistic MIMO U-Net

The best solution of the Weather Prediction track in the Yandex Shifts challenge

[ICLR 2023] Deep Ranking Ensembles for Hyperparameter Optimization

An Ensemble Deep Learning Model to Classify Scoliosis and Healthy Subjects Based on Non-invasive Rasterstereography Analysis

[ICML 2024] Boundary Exploration for Bayesian Optimization With Unknown Physical Constraints

Full-Reference Image Quality Assessment models based on ensemble of gradient boosting

early detection method for parkinson's disease using deep ensemble learning on MRI dataset

Companion code for the MICCAI 2026 paper “Lost in the Folds” for uncertainty estimation in medical image segmentation.

Ransomware analysis using DEL with jet-like architecture comprising two CNN wings, a sparse AE tail, a non-linear PCA to produce a diverse feature space, and an MLP nose

resVAE-ensemble: feature identification in single-cell data

Physics-informed neural surrogate for structural analysis — 1000x faster than FEA with >99.9% accuracy

A pedagogical study and analysis of the results from "Deep Ensembles: A Loss Landscape Perspective".

Kvantifikacija nesigurnosti u modelima umjetne inteligencije: okvir za prediktivno održavanje i analizu rizika

Reproducible pipeline for the integrated downscaling–BMA–BiLSTM framework: CMIP6 GCM evaluation, two-track downscaling (LARS-WG 8 + DQM), EM-based Bayesian Model Averaging, deep-ensemble streamflow projection, and drought / ETCCDI / compound-extreme analysis. Demonstrated on the Haraz basin, northern Iran.

Uncertainty-aware single-channel sEMG hand gesture classification with MC Dropout selective prediction — LOSOCV benchmark on YTU dataset and NinaPro DB2 external validation

RUL prediction for turbofan engines on NASA N-CMAPSS (DS02): classical ML (NGBoost + conformal prediction) vs. deep learning (CNN deep ensemble) with calibrated uncertainty quantification.