https://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=CC82E9E0A5AA81DC8163412A904747D3?doi=10.1.1.99.9681&rep=rep1&type=pdf

Semi-Supervised Learning Literature Survey

D, F. (2017, October 11). Simple explanation of Semi-Supervised Learning and Pseudo Labeling. Retrieved March 14, 2020, from [towardsdatascience.com/simple-explanation-of-semi-supervised-learning-and-pseudo-labeling-c2218e8c769b](https://towardsdatascience.com/simple-explanation-of-semi-supervised-learning-and-pseudo-labeling-c2218e8c769b) 

Simple explanation of Semi-Supervised Learning and Pseudo Labeling (Reference)

Pseudo labeling is a process used in semi-supervised learning, where it combines both supervised and unsupervised techniques in order to train a model. The algorithm is as follows:
1. You use the portion of your data set that is already labeled as the training set to train your model (supervised learning)
2. Once the model is giving you good results, you then use that model with your unlabeled data (unsupervised learning)
    - The model will predict the outcomes of the unlabeled data. We do not know if these are correct, but we have accurate labels after training it with our labeled data
3. We then combine both the original training set with the unlabeled data that was just labeled, training our model on the full data set
Pseudo labeling helps improve the model as well as reduce the error.

Pseudo Labeling

Semi-supervised learning lies somewhere between supervised and unsupervised learning. It takes small amount of labeled data to start learning and then moves on to huge quantity of unlabaled data to develop further.

University of Michigan - Ann Arbor

- Supervised statistical learning
- Unsupervised statistical learning
- Time series learning
- Semi-Supervised learning

Types of statistical learning

Unsupervised machine learning methods find structure in unlabeled data.
They are used when we DON'T have prior knowledge about the categories.

Unsupervised statistical learning

Semi Supervised learning

Differences about the Supervised vs Unsupervised Machine learning

Training a model using labeled data and using this model to predict the labels for new data is known as __________.

Modeling the features of an unlabeled dataset to find hidden structure is known as _____.

A time series is a set of measurements of a quantitative variable taken at sequential intervals over a continuous period of time. This approach allows researchers to track trends and ongoing changes within a specific domain, such as recording a factory's weekly worker productivity over an entire year.

Time Series

Supervised learning is a type of machine learning problem where the objective is to predict a designated unknown label based on known inputs. This approach requires a training dataset consisting of examples for which the ground truth labels are already known, allowing the model to learn the mapping from inputs to classifications.

Learn Before

Related

Learn After