Think of this as the one-stop-shop/dictionary/directory for your machine learning algorithms. The algorithms have been sorted into 9 groups: Anomaly Detection, Association Rule Learning, Classification, Clustering, Dimensional Reduction, Ensemble, Neural Networks, Regression, Regularization. In this post, you’ll find 101 machine learning algorithms, including useful infographics to help you know when to use each one (if available).
101 Machine Learning Algorithms
Each of the accordian drop downs are embeddable if you want to take them with you. All you have to do is click the little ’embed’ button in the lower left hand corner and copy/paste the iframe. All we ask is you link back to this post.
By the way, if you have trouble with Medium/TDS, just throw your browser into incognito mode.
Any of these classification algorithms can be used to build a model that predicts the outcome class for a given dataset. The datasets can come from a variety of domains. Depending upon the dimensionality of the dataset, the attribute types, sparsity, and missing values, etc., one algorithm might give better predictive accuracy than most others. Let’s briefly discuss these algorithms. (18)
Regression Analysis is a statistical method for examining the relationship between two or more variables. There are many different types of Regression analysis, of which a few algorithms can be found below. (20)
A neural network is an artificial model based on the human brain. These systems learn tasks by example without being told any specific rules. (11)
Also known as outlier detection, anomaly detection is used to find rare occurrences or suspicious events in your data. The outliers typically point to a problem or rare event. (5)
With some problems, especially classification, there can be so many variables, or features, that it is difficult to visualize your data. Correlation amongst your features creates redundancies, and that’s where dimensionality reduction comes in. Dimensionality Reduction reduces the number of random variables you’re working with. (17)
Ensemble learning methods are meta-algorithms that combine several machine learning methods into a single predictive model to increase the overall performance. (11)
In supervised learning, we know the labels of the data points and their distribution. However, the labels may not always be known. Clustering is the practice of assigning labels to unlabeled data using the patterns that exist in it. Clustering can either be semi parametric or probabilistic. (14)
Association Rule Analysis
Association rule analysis is a technique to uncover how items are associated with each other. (2)
Regularization is used to prevent overfitting. Overfitting means the a machine learning algorithm has fit the data set too strongly such that it has a high accuracy in it but does not perform well on unseen data. (3)
Scikit-Learn Algorithm Cheat Sheet
First and foremost is the Scikit-Learn cheat sheet. If you click the image, you’ll be taken to the same graphic except it will be interactive. We suggest saving this site as it makes remembering the algorithms, and when best to use them, incredibly simple and easy.
SAS: The Machine Learning Algorithm Cheat Sheet
You can also find many of the same algorithms on SAS’s machine learning cheet sheet as the one above. The SAS website (click the pic) also gives great descriptions about how, when, and why to use each algorithm.
Microsoft Azure Machine Learning: Algorithm Cheat Sheet
Microsoft Azure’s cheet sheet is the simplest cheet sheet by far. Even though it is simple, Microsoft was still able to pack a ton of information into it. Microsoft also made their algorithm sheet available to download.
There you have it, 101 machine learning algorithms with cheat sheets, descriptions, and tutorials! We hope you are able to make good use of this list. If there are any algorithms that you think should be added, go ahead and leave a comment with the algorithm and a link to a tutorial. Thanks!