All Research
Machine Learning
From perceptrons to transformers - neural architectures, optimization, loss functions, and representation learning.
Neural NetworksBackpropagationOptimizationLoss FunctionsAttention MechanismsEmbeddingsRepresentation LearningTransformers
Articles
2 publications in machine learning
·ml
Spectral Encoders for Signal Classification
Designing neural network architectures that learn directly from spectral representations. Our encoder maps frequency-domain features into a latent space optimized for structural classification.
neural networksspectral learningencoders
·ml
Anomaly Detection in Resonance Patterns
Using variational autoencoders to detect anomalous resonance patterns in structural health monitoring. A signal-based approach to predictive maintenance.
anomaly detectionVAEresonance