Classification Algorithm Overview

RestNet-50

A Convolutional neural network that is 50 layers deep. You can load a pre-trained version of the network that has been trained on more than one million images from the ImageNet database. The pre-trained network can classify images into 1000 object categories, such as a keyboard, mouse, pencil, and many animals.

Convolutional Neural Network

A class of artificial neural network commonly applied to the analysis of images, also known as a “ConvNet”, “shift invariant” or “SIANN”

Transfer Learning

Focused on storing knowledge gained while solving one problem and applying it to a different, but related, problem. For example, knowledge gained while learning to recognize cars could apply when trying to recognize trucks. Technically this is fine tuning of a network that was pre-trained on some big dataset with new classification layers - reduces training time without sacrificing accuracy. Achieved by using “bottlenecks”, the internal representation of one of the input samples in the network, at a certain depth level.

Tensorflow

An open-source library for machine learning natively supported in Kubeflow - excellent for the training and inference of deep neural networks.