Optuna vs. Grid Search: Hyperparameter Optimization at Scale for Deep Learning
After running 1,200 trials across three medical imaging models, here's what I learned about Bayesian optimization, pruning strategies, and when to stop searching.
About the author
Vanshaj Awasthi is a Full Stack Developer in Pune, India, with 17+ IEEE and Springer publications in medical AI. See the full publication list.
More writing
- From EEG Noise to Neural Commands: Building a GenAI-Enhanced BCI
Our ICEIL 2024 paper in plain English — how GAN-based data augmentation turned noisy EEG signals into reliable communication…
- Deploying ML Models on AWS Lambda: A Practical Guide for Researchers
Everything I wish I knew before containerizing PyTorch models for serverless inference — cold starts, ONNX Runtime, layer…