{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# BASIC SEMI-SUPERVISED LEARNING MODELS" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Sometimes pressing Enter in TensorFlow is not enough to get good results. Sometimes you have too little data to work with and maximum likelihood estimation gives you some bizarre answer. Or even more common - data and labeling are *expensive* and the relationship between the number of instances in your dataset and quality of the final model is not linear at all.\n", "That's when semi-supervised learning (SSL) come to play.\n", " \n", " > Semi-supervised learning is a class of machine learning tasks and techniques that also make use of unlabeled data for training – typically a small amount of labeled data with a large amount of unlabeled data [[source]](https://en.wikipedia.org/wiki/Semi-supervised_learning).\n", " \n", "These types of ML algorithms are using both labeled and unlabeled data to get more accurate predictions (or in rare cases to get them at all). " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "