Machine Learning is a field of study focused on computer algorithms designed to perform tasks without being explicitly programmed to do so. Another interpretation is to consider machine learning algorithms as those able to improve in performance on a specific task, assessed by a specific measure, after experience with said task.
Machine learning is a subset within the broader field of Artificial Intelligence and commonly used in fields such as data science, bioinformatics, natural language processing, and computer vision. Other common consumer-facing applications include product recommendation engines, price comparison tools, and image recognition software.
Types of Machine Learning
The two most common types of machine learning are supervised and unsupervised machine learning. These types of ML algorithms address a wide range of applications ranging from predicting housing prices to risk factors of specific diseases. While broad in scope, they differ quite significantly in their foundational design.
Supervised Learning
Supervised machine learning algorithms use training data where the desired output is given. For example, a set of features on new cars paired with the final sale price—the price being the given outcome. The machine learning model would be trained on this data to predict the market price for data where the car features were given but the price was unknown.
Unsupervised Learning
Supervised machine learning models are designed to work with datasets that have no outcome variable and seek only to discover structure within the data. Common applications include clustering, which has many applications within fields such as medicine, finance, and statistical modeling. For example, genetics researchers may group certain individuals together based on their specific genetic expressions.
Other Types of Machine Learning
Supervised and Unsupervised are merely the most popular types of machine learning algorithms, particularly those that see widespread industry use (R). An incomplete listing of these types of machine learning are as follows:
- Semi-Supervised Learning
- Reinforcement Learning
- Feature Learning
- Sparse Dictionary Learning
- Anomaly Detection
Machine Learning Algorithms
In addition to the many types of machine learning, there are also types of machine learning algorithms. These may be used among several different types of models or be more restricted to use among a single model (such as supervised learning.) A list of the most common machine learning algorithms is as follows:
- Support Vector Machines
- Naive Bayes
- Random Forest
- Linear Regression
- Logistic Regression
- Decision Tree
- K-Means
