Library · Machine learning & AI
What Transfer Learning & Domain Adaptation got wrong, from 48 theses
Summary written by a model from the records below: The records document numerous failures and baseline losses across transfer learning and domain adaptation techniques in diverse engineering, scientific, and vision tasks. Common failure points include negative transfer across distinct domains, poor performance relative to standard non-transfer baselines, and ineffective distribution alignment. These records come from graduate theses, 2021 to 2026, each labeled with its degree where the repository states it, and each links to its thesis. They were extracted by language models reading the full text, so treat each as a lead to read, not a verdict.
Plus 9 problems the authors left open, listed at the end.
Failure modes, by theses reporting them
Select a mode to jump to its records.
Transfer learning underperforms target-only training and simple non-transfer baselines
Transfer learning models frequently failed to improve upon or were beaten by training directly on the target dataset, naive baselines, or random weight initialization. When task similarity was low or target datasets had distinct properties, transferring source representations caused negative transfer or provided no convergence benefits.
Tried and failed
transfer learning via large-scale pretraining applied to industrial semantic segmentation. Outcome: worse than baseline. Reason: pretraining on related domain datasets underperformed training from scratch on the target domain
Beyond standard benchmarking : towards robust and trustworthy robotics for industrial and nuclear applications · UT Austin · PhD
Lost to a baseline
Transfer learning on Goodfellow CNN (10s) on ICU test set achieved F1=0.85, matching zero-shot ECG-FM v1 (F1=0.85) and beaten by non-transfer Inception v3 recurrence plots (F1=0.88).
Atrial Fibrillation Detection from Electrocardiograms of Intensive Care Unit Patients: A Comparison of Artificial Intelligence Approaches · Queen's University · Master's
Lost to a baseline
Zero-information transfer learning (testing source classifiers directly on target memory data with simple score averaging) performed significantly worse than typical unidimensional memory classification (t(42) = 1.90, p = 0.032)
More than sum of its parts : investigating episodic memory as a multidimensional cognitive process · UT Austin · PhD
Lost to a baseline
At very low task similarity rho in transfer learning, standard learning with no transfer (delta = 0) beats both hard and soft transfer.
Estimation and Learning via Convex Optimization: Asymptotics, Phase Transitions, and New Algorithms · Harvard · PhD
Lost to a baseline
Transfer learning models for zinc(II) salphen emission energy (Fingerprint + Ridge R2 = 0.831 ± 0.095, MAE = 0.0769 ± 0.0216 eV) showed no significant improvement over the baseline model trained purely on the 49 zinc(II) salphen dataset (Fingerprint + Ridge R2 = 0.825 ± 0.216, MAE = 0.0706 ± 0.0429 eV)
Discovery of photoactive anti-cancer platinum(II) complexes: synthesis, screening and machine learning approaches · Imperial · PhD
Considered and rejected
Considered and rejected: Rejected pure transfer learning due to decreased classification performance compared to retraining dataset-specific models.
Automated Design and Optimization of Metallic Alloys · University of Nevada, Reno · Master's
Tried and failed
transfer learning from lower-fidelity pre-trained neural networks applied to neural network force field training. Outcome: worse than baseline. Reason: pre-trained weights provided no faster convergence or accuracy gain over random initialization
Machine-learning models for analysis of biomass reactions and prediction of reaction energies · Georgia Tech · PhD
Lost to a baseline
Model-free tracking MSE on target SISO system after transfer learning (2.135e-5) was slightly worse than model-based control with an accurate predictor.
Control of Agentic Systems Using the Newton-Raphson Controller · Georgia Tech · PhD
Lost to a baseline
Transfer learning on 2 clusters (MAE 0.127) performed worse than the K=1 base model (MAE 0.123) for LSTM on the 4-week cell-A sample.
A Framework for Generalizing Uncertainty in Mobile Network Traffic Prediction · Virginia Tech · Master's
Lost to a baseline
Smaller fully-connected neural network models (e.g., 5-2 and 5-5) performed worse when trained with transfer learning (train RMSE 2.80 mm and 2.49 mm) than when trained only with self-supervised learning (2.20 mm and 1.72 mm) or supervised learning (0.97 mm and 1.48 mm).
Enabling Shape-Based Approaches for Autonomous Percutaneous Interventions with Sensorized Needles · Johns Hopkins · PhD
Lost to a baseline
Multimodal transfer learning flood prediction achieved 0.783 accuracy (1-year), lost to naive persistent baseline (0.895 accuracy).
Multimodal Machine Learning for Climate Adaptation · MIT · PhD
Lost to a baseline
Cold-started and warm-started transfer learning CNN models on M. buryatense copper response performed no better than null baseline models trained on shuffled sequences (F1 ~0.34 vs shuffled F1 ~0.33; Pearson r ~0.26 vs shuffled r ~0.19).
Towards sustainable biomolecule production: computational approaches to accelerate genetic tool development for engineering metabolism in microorganisms · University of Washington · PhD
Lost to a baseline
Transfer learning models for platinum(II) NNC emission energy (Mordred + RF R2 = 0.760 ± 0.123, test R2 = 0.18, test MAE = 0.14 eV) showed no significant improvement over the baseline model trained purely on the platinum(II) NNC dataset (Mordred + RF R2 = 0.723 ± 0.241, MAE = 0.0927 ± 0.0279 eV)
Discovery of photoactive anti-cancer platinum(II) complexes: synthesis, screening and machine learning approaches · Imperial · PhD
Domain adaptation alignment methods fail to improve over standard baselines
Specialized unsupervised and adversarial domain adaptation methods failed to outperform simple empirical risk minimization, data merging, or unadapted models. These alignment approaches suffered from poor target label estimation, degraded transferability in deeper backbones, or an inability to guarantee class-level feature alignment.
Tried and failed
unsupervised domain adaptation with distribution alignment applied to large-scale image classification. Outcome: worse than baseline. Reason: negative transfer occurred due to poor target label estimation when source data was plentiful
TOWARDS ROBUST VISUAL PERCEPTION SYSTEMS IN REAL-WORLD ENVIRONMENTS · Cornell · PhD
Tried and failed
domain adaptation and generalization algorithms applied to tabular neuroimaging classification across sites. Outcome: worse than baseline. Reason: Specialized domain adaptation methods failed to significantly outperform standard empirical risk minimization.
Fair and Generalizable Machine Learning for Neuroimaging · Penn · PhD
Lost to a baseline
Multi-source domain adaptation methods were outperformed by simple merging on the Heintz-Buschart real-data target domain.
Statistical Methods for Outcome Measurement Error Correction, And Multi-study Prediction And Causal Inference under Study Heterogeneity · Harvard · PhD
Tried and failed
adversarial feature and output alignment applied to LiDAR depth completion domain adaptation. Outcome: no signal. Reason: provided minimal performance impact and failed to bridge the domain gap
Domain adaptation for semantic and 3D tasks · Imperial · PhD
Tried and failed
increasing backbone neural network depth applied to unsupervised domain adaptation. Outcome: did not generalise. Reason: deeper networks did not resolve domain discrepancy and slightly reduced feature transferability across domains
Domain Adaptation with a Classifier Trained by Robust Pseudo-Labels · Virginia Tech · Master's
Tried and failed
unsupervised domain adaptation for multiple instance learning applied to out-of-distribution histopathology slide classification. Outcome: did not generalise. Reason: feature and pixel adaptation failed to provide statistically significant improvements on distinct external target cohort
Domain Adaptation for Breast Cancer Computational Pathology: Evaluating Feature-Level and Pixel-Level Approaches Under Leave-One-Domain-Out Framework · Harvard · Master's
Lost to a baseline
Real-time unsupervised domain adaptation performed statistically worse than the semi-supervised baseline on level ground walking R2 (0.64 vs 0.77).
Enabling Scalable, Versatile, and Robust Control for Robotic Exoskeletons · Georgia Tech · PhD
Lost to a baseline
Unsupervised domain adaptation alone using CycleGAN without BWE (11.50% EER / 0.532 minDCF on SRE16-YUE-eval40) degraded performance relative to the unadapted/no-BWE baseline (7.46% EER / 0.382 minDCF).
Robust Speaker Recognition using Perceptual and Adversarial Speech Enhancement · Johns Hopkins · PhD
Considered and rejected
Considered and rejected: Adversarial domain adaptation was rejected in favor of pseudo-shot learning because adversarial training aligns only domain-level distributions without ensuring correct class-level feature alignment
Coping with the distribution change in soil classification with LIBS · University of Regina · Master's
Considered and rejected
Considered and rejected: Rejected transfer component analysis (TCA) for domain adaptation because it extracts representations without utilizing source domain labels, which can discard features with moderate domain mismatch but high predictive power
Towards Robust Machine Learning for Health Applications · Tübingen · PhD
Severe distribution shift across environments or data modalities prevents generalization
Direct transfer across substantially different acoustic environments, biological systems, sensor resolutions, or operational aging regimes failed to generalize. Discrepancies such as differences between in vitro cell lines and clinical tumors led to severe performance degradation and negative transfer.
Tried and failed
transfer learning with synthetic pretraining and finetuning applied to subsurface seismic fault segmentation. Outcome: did not generalise. Reason: negative transfer and inductive bias mismatch between distinct synthetic seismic geological models
Multiscale Integration of Cross-Modal Subsurface Data for Reservoir Characterization under Label-Constrained Environments · Georgia Tech · PhD
Tried and failed
2D transfer learning segmentation on 3D volumes applied to low-resolution medical image segmentation. Outcome: did not generalise. Reason: Reduced signal and resolution caused severe sensitivity loss for small target regions below a size threshold.
Machine Learning And Quantitative Neuroimaging In Epilepsy And Low Field Mri · Penn · PhD
Tried and failed
zero-shot transfer learning with pretrained audio embeddings applied to idiosyncratic atypical vocalization classification. Outcome: no signal. Reason: generic pretrained representations failed to capture idiosyncratic acoustic patterns of atypical vocalizations
Foundations of Cognitive, Affective, and Communicative Systems for Neurodiverse Individuals · MIT · PhD
Tried and failed
cross-region transfer learning for acoustic classification applied to audio event detection. Outcome: did not generalise. Reason: models trained on external or online datasets failed to transfer to a new geographic acoustic environment
Listening in on the forest: use of bioacoustics to preserve soundscapes and rare species · Imperial · PhD
Lost to a baseline
Generic zero-shot AudioSet transfer learning achieved only 51.1% accuracy on self-talk, performing barely above chance
Tried and failed
transfer learning for long-term degradation prediction applied to battery state of health estimation. Outcome: did not generalise. Reason: model failed to extrapolate capacity loss during late-life aging phases beyond trained cycle horizons
Modeling and Simulation of Power System with High Penetration of Inverter-based Resources · Georgia Tech · PhD
Tried and failed
direct cross-dataset model transfer without domain adaptation applied to cross-study drug sensitivity prediction. Outcome: did not generalise. Reason: distribution mismatch between source and target datasets degraded predictive performance without domain mapping
Application of advanced machine learning based approaches in cancer precision medicine · Texas Tech · PhD
Tried and failed
direct transfer learning from preclinical cell lines applied to clinical patient drug response prediction. Outcome: did not generalise. Reason: distribution shift and biological discrepancies between in vitro cell models and in vivo human tumors
Predicting cancer patient response to chemotherapy using machine learning from small data · Imperial · PhD
Tried and failed
transfer learning across related domains applied to polygenic risk score prediction. Outcome: did not generalise. Reason: insufficient shared genetic architecture between auxiliary and target traits
TRANSFER LEARNING IN CLASSIFICATION AND REGRESSION WITH SUMMARY STATISTICS · Penn · PhD
Lost to a baseline
Transfer learning of eqt-pnw was outperformed by semblance ensembling in cross-domain picking accuracy on noisy data
Observing Seismic Variations by Earth and Lab Fluids and Fractures · Harvard · PhD
Pretrained visual representations fail on specialized and non-standard image modalities
Pretrained natural image representations transferred poorly to specialized domains including fluorescence microscopy, radio frequency spectrograms, malware graphs, and thermal images. These models suffered from pretraining biases, severe overfitting, or worse performance than training from random initialization.
Tried and failed
transfer learning for image segmentation applied to noisy infrared thermal images. Outcome: worse than baseline. Reason: struggled to segment low signal-to-noise regions compared to hybrid fuzzy clustering
Autonomous Experimentation to Accelerate Boiling Heat Transfer Research · MIT · PhD
Tried and failed
transfer learning with natural image pretrained weights applied to satellite land cover image segmentation. Reason: None
Practical methods for the advancement of precision conservation via land cover classification and conformal prediction · Iowa State · PhD
Tried and failed
ImageNet pre-trained transfer learning applied to calcium imaging neural decoding. Outcome: worse than baseline. Reason: natural image visual representations transfer poorly to fluorescence microscopy signals compared to training from scratch
Optimizing sensorimotor behaviors through information integration and mental simulation · MIT · PhD
Tried and failed
transfer learning with natural image pretraining applied to malware graph visual representation classification. Outcome: worse than baseline. Reason: None
Developing Robust Models, Algorithms, Databases and Tools With Applications to Cybersecurity and Healthcare · Georgia Tech · PhD
Tried and failed
transfer learning with natural image pre-trained models applied to simulated microscopy image classification. Outcome: worse than baseline. Reason: Pre-training bias hindered adaptation compared to training from random initialization
Mitigating effects of data heterogeneity in decentralized learning and addressing challenges of adapting foundation models · Iowa State · PhD
Tried and failed
transfer learning with natural image pre-trained weights applied to RF spectral eigengram object detection. Outcome: worse than baseline. Reason: natural image visual features did not transfer well compared to random initialization for non-visual spectrogram representations
Intelligently Leveraging Multi-Channel Image Processing Neural Networks for Multi-View Co-Channel Signal Detection · Virginia Tech · Master's
Tried and failed
ImageNet pre-trained transfer learning applied to grayscale image classification. Outcome: did not generalise. Reason: Natural image features do not transfer effectively to grayscale, manufacturing, or medical domains
Engineering-Driven Learning Approaches for Bio-Manufacturing and Personalized Medicine · Georgia Tech · PhD
Tried and failed
transfer learning with pre-trained convolutional neural networks applied to droplet deposition density map classification. Outcome: overfit. Reason: models underperformed severely on test data and suffered from poor generalization and overfitting
Automated Exploration of High-Mix, Low Volume Direct Write Design Spaces Through Artificial Intelligence · Georgia Tech · PhD
Domain adaptation provides no benefit under insufficient or non-overlapping discrepancy
Domain adaptation algorithms yielded no measurable improvement when source and target distributions lacked sufficient domain shift or matched uniform priors. Similarly, transferring between domains failed when feature distributions such as synthetic signal-to-noise ratios had non-overlapping ranges.
Tried and failed
unsupervised domain adaptation with adversarial training applied to intra-position gesture classification. Reason: source and target distributions within identical positions lacked sufficient domain shift for adaptation benefits
Adaptive gesture recognition for human-robot interface using mechanomyography (MMG) · Imperial · PhD
Lost to a baseline
Transferring from easier (high SNR) to harder (low SNR) synthetic domains was beaten by the limited-data baseline when SNR ranges did not overlap
Foundations of Radio Frequency Transfer Learning · Virginia Tech · Doctorate
Tried and failed
unsupervised domain adaptation applied to tasks with small domain discrepancy. Outcome: worse than baseline. Reason: source and target domains already had minimal discrepancy, rendering adaptation alignment ineffective
Overcoming Distribution Shifts and Imbalance Challenges in Representation Learning for Deep Regression Models · EPFL · PhD
Tried and failed
learning prior class distribution in optimal transport applied to unsupervised domain adaptation. Outcome: no signal. Reason: the target domain had a balanced class distribution matching a uniform prior
A prototype-oriented framework for deep transfer learning applications · UT Austin · PhD
Freezing feature layers and rigid transfer constraints degrade target adaptation
Freezing early neural network layers or relying on fixed-feature transfer degraded target predictive performance compared to full network fine-tuning. These rigid representation constraints failed to capture nonlinear feature interactions and could not mitigate the transfer of representation bias.
Tried and failed
transfer learning with frozen feature layers applied to time series classification. Outcome: worse than baseline. Reason: None
Statistical Machine Learning on Time Series with Applications to Manufacturing and Healthcare · Texas Tech · PhD
Considered and rejected
Considered and rejected: Transfer learning by freezing/constraining early neural network layers was rejected because it degraded prediction performance relative to full-network fine-tuning.
Building Blocks of Neural Network Intermolecular Interaction Potentials · Georgia Tech · PhD
Tried and failed
transferring linear representation projections to nonlinear kernels applied to molecular property prediction. Outcome: did not generalise. Reason: linear optimization does not account for higher-order feature interactions introduced by polynomial kernel powers
A general and efficient framework for atomistic machine learning · EPFL · PhD
Tried and failed
weight decay regularization applied to fixed-feature transfer learning. Outcome: did not generalise. Reason: does not mitigate transfer of representation bias
Towards ML Models That We Can Deploy Confidently · MIT · PhD
Left open by the authors
Problems the authors named and did not get to.
Left open
Apply formal transfer learning algorithms to measure and mitigate negative transfer between simulated genomic data and empirical target datasets. Blocker: None
The Inference of Selective Sweep Parameters from their Genomic Footprint · Cornell · PhD
Left open
Scale the multi-step Vision Transformer transfer learning framework to broader medical image classification and segmentation tasks across modalities. Blocker: None
Detecting Covid-19 Effectively With Transformers And CNN-based Deep Learning Mechanisms · Texas Tech · Master's
Left open
Implement transfer learning using pre-trained computer vision models on low-light plant image datasets for leaf segmentation. Blocker: Requires the private low-light and EMCCD luminescence plant imaging dataset from the thesis lab
Automation of Luminescence Quantitation for High-Throughput Plant Phenotyping Using Image Processing and Deep Learning · Texas State University · Master's
Left open
Develop transfer learning and domain adaptation methods to relax transportability assumptions between clinical trial surrogate marker datasets. Blocker: No specific statistical framework or target algorithm is defined beyond general concepts
Heterogeneous surrogate markers in clinical trials and real-world settings · UT Austin · PhD
Left open
Develop domain adaptation algorithms combined with spectral data calibration for LIBS soil classification across distribution shifts. Blocker: Access to the thesis's specific calibrated LIBS soil spectral datasets.
Coping with the distribution change in soil classification with LIBS · University of Regina · Master's
Left open
Analyze theoretical and empirical conditions under which simple multi-source data merging outperforms domain adaptation in regression settings. Blocker: None
Statistical Methods for Outcome Measurement Error Correction, And Multi-study Prediction And Causal Inference under Study Heterogeneity · Harvard · PhD
Left open
Develop few-shot domain adaptation methods to adapt large pre-trained models to novel domains with limited annotations. Blocker: None
Learning with Limited Labeled Data: Techniques and Applications · Virginia Tech · Doctorate
Left open
Investigate domain adaptation methods for short-sequence LSTM ensembles applied to significantly different source and target dynamic systems. Blocker: The objective is a broad research direction without specific target datasets, adaptation approaches, or evaluation metrics.
Parameter optimization for enhanced modeling of dynamic systems · Iowa State · PhD
Left open
Develop unified time-series foundation models across diverse datasets using transfer learning and domain adaptation techniques. Blocker: None
Graph-based Time-series Forecasting in Deep Learning · Virginia Tech · Doctorate
Who has worked on this
Authors of the public theses cited on this page, most-cited first.
- Molly Bartlett · Imperial · PhD · 2025
Discovery of photoactive anti-cancer platinum(II) complexes: synthesis, screening and machine learning approaches - Yujie Wu · Harvard · PhD · 2024
Statistical Methods for Outcome Measurement Error Correction, And Multi-study Prediction And Causal Inference under Study Heterogeneity - Yingchao Huang · University of Regina · Master's · 2024
Coping with the distribution change in soil classification with LIBS - Jiarong Ye · Harvard · Master's · 2026
Domain Adaptation for Breast Cancer Computational Pathology: Evaluating Feature-Level and Pixel-Level Approaches Under Leave-One-Domain-Out Framework - Hanqin Du · Imperial · PhD · 2026
Predicting cancer patient response to chemotherapy using machine learning from small data - Kaicheng Niu · Georgia Tech · PhD · 2025
Control of Agentic Systems Using the Newton-Raphson Controller - Sarah Belal Nassar · Queen's University · Master's · 2025
Atrial Fibrillation Detection from Electrocardiograms of Intensive Care Unit Patients: A Comparison of Artificial Intelligence Approaches - Ismail Nejjar · EPFL · PhD · 2025
Overcoming Distribution Shifts and Imbalance Challenges in Representation Learning for Deep Regression Models - Siyao Cai · Georgia Tech · PhD · 2024
Modeling and Simulation of Power System with High Penetration of Inverter-based Resources - Congcong Yuan · Harvard · PhD · 2024
Observing Seismic Variations by Earth and Lab Fluids and Fractures - Cynthia Zeng · MIT · PhD · 2024
Multimodal Machine Learning for Climate Adaptation - Madhumitha Ravichandran · MIT · PhD · 2024
Autonomous Experimentation to Accelerate Boiling Heat Transfer Research
Checking a claim in this area?
We can run the same search on any method or claim. If nothing turns up, we will say so, and that proves nothing on its own.