AI4ALL: Basics in Artificial Neural Network

Model progress can be saved during and after training. This means a model can resume where it left off and avoid long training times. Saving also means you can share your model and others can recreate your work. When publishing research models and techniques, most machine learning practitioners share:

• code to create the model, and

• the trained weights, or parameters, for the model

Converts a class vector (integers) to binary class matrix.

Regularizers allow you to apply penalties on layer parameters or layer activity during optimization. These penalties are summed into the loss function that the network optimizes.

Just your regular densely-connected NN layer.

Loss functions measure the amount of mistakes our model is making. It’s a crucial component in neural networks because it is the objective function for the optimizer.

Optimizer is a choice of different optimization techniques. It is a tuning parameter because each optimizer converges at different speed.

The compiler puts together optimizer and the loss function so that the neural network is ready for training.

Training the neural network can be done using .fit method. The number epochs and batch sizes are tuning parameter.

The evaluation is important because we need to separate training and testing data. The evaluation is almost always done on test data. Test data is the data that has not been appeared in training data.

Save your model so that you do not lose your work progress.

Create an installable git package so you can save your production code for deployment.