Skip to content

binary classifier

A single artificial neuron – I – a primitive ANN for a classification problem

This post requires Javascript to display formulas! When you start working with Artificial Neural Networks [ANNs] there are a lot of things you must get familiar with: Different types of networks and network layers, weights, signal propagation, loss, backward error propagation, gradient descent, regularization, normalization, tensors (arrays) …. In addition you may have to fight with complex layer structures even for relatively simple experiments. And the objects you work with will typically be described in high-dimensional variable spaces. The good news is that one can… Read More »A single artificial neuron – I – a primitive ANN for a classification problem