TensorFlow

Sequential

Sequential is a keras class for creating a linear stack of layers. The stack here reflects both the conceptual and technical design of how this Keras class works. 1. Linear, Ordered Structure A stack implies a strict, sequential order where layers are piled one on top of the other. 2. In dl, we often think of networks as having depth, with layers stacked vertically in diagrams.

Faltten

A keras layer that transforms a multi-dimensianl input into a 1D vector. It collapses all dimensions except the batch size into a single dimension . It's a purely a format change - it reshapes the data without any mathematical calculations.

Dense

A fully connected layer in Keras where every input neuron connects to every output neuron. Each of the 50 neurons computes a weighted sum of all 832 inputs, adds a bias, then applies ReLU.

model.compile

A keras method that configures the model for training by specifying how it will learn and evaluate performance.