Machine Learning Fields
Machine Learning algorithms are usually classified into 3 to 4 groups.
The first group contains algorithms which learn the relation between given input-output data pairs. Because the input as well as the output (target) values are given, this group is called Supervised Learning.
The next group of algorithms are used to find the underlying structure of data without telling the algorithm what it has to find. For this reason this group is called Unsupervised Learning.
The third group is somehow a mixture of the properties of the first two and is therefore called Semi-Supervised Learning. Another main group of machine learning algorithms is called Reinforcement Learning.
These algorithms solve problems by exploring an environment and learning from - sometimes very sparse - returned reward which can be either positive or negative.
Supervised Learning
Building a model from labeled data or input-output data pairs.
Regression
- Neural Networks
Classification
- Neural Networks
- K Nearest Neighbors (KNN)
- Support Vector Machines (SVM)
- Decision Trees
- Random Forests
- Naive Bayes
Time Series Prediction
- Neural Networks
- Recurrent Neural Networks (RNN)
- Long Short Term Memory (LSTM)
- Gated Recurrent Unit (GRU)
Unsupervised Learning
Learning the underlying structure of data without given labels.
Clustering
- K-Means
Dimensionality Reduction
- Principal Component Analysis (PCA)
Semi-Supervised Learning
Learning the underlying principal feature distribution and generating new but similar data.
Generative Models
Reinforcement Learning
Solving tasks by exploring an environment and evaluating the returned reward.