A survey of neuromorphic hardware and algorithm advances with guidance on future directions.
Using machine learning and computer vision to automate the diagnosis of diabetic retinopathy from clinical eye images.
Using deep convolutional neural networks, transfer learning, and fit one cycle optimisations to achieve 98.6% accuracy on detecting cancer in the PatchCamelyon dataset.
Guided walkthrough of neural network architecture, activations, forward passes, softmax outputs, losses, and sensible parameter initialization.
Guided walkthrough of logistic regression covering optimisation, regularisation (L1/L2), threshold tuning, and calibration diagnostics.
Guided walkthrough of univariate linear regression: derive gradient descent, tune learning rates, and validate convergence with code.