Logistic Regression is a supervised machine learning algorithm used to predict the probability of an outcome belonging to a particular category. Unlike ordinary regression, which is generally used to predict numerical values, Logistic Regression is designed for...

Ridge Regression is a machine learning algorithm used to predict numerical outcomes while assisting with situations where input variables are strongly related to one another. This can be a problem in traditional regression due to the highly correlated variables that...

It's often difficult to pinpoint the parts of the data which are most useful when analyzing datasets. A machine learning technique that simplifies this issue by identifying important rows and columns within a dataset is CUR Matrix Decomposition. Like in Oracle ML, it...

Oftentimes, it isn’t very obvious how the different records in datasets should be grouped. There may be patterns inside the data, but no predefined categories to tell us where each record belongs. Expectation Maximization or EM for short, is an unsupervised machine...

Not all machine learning problems are about predicting a known outcome. Sometimes the more interesting question is whether something looks unusual when compared with everything else. This is the idea behind One Class Support Vector Machine (One Class SVM). Supported...

Large datasets can contain a huge amount of information, this does not always make analysis easier. In many cases, the challenge is finding a simpler way to represent the data while keeping the patterns that actually matter. Non Negative Matrix Factorization (NMF) is...

Finding significant data groups within datasets can be a challenge, especially when there are no predefined categories to work with. O Cluster, short for Orthogonal Partitioning Clustering, is an unsupervised machine learning algorithm developed by Oracle to...

Understanding text can be a difficult task for computers. While people can easily recognize the meaning behind words and phrases, computers need methods that assists them interpret language in a meaningful way. Explicit Semantic Analysis, also referred to as ESA, is a...

Singular value decomposition is an unsupervised machine learning technique used primarily for feature extraction and the transformation of high dimensional data into a lower dimensional space whilst keeping meaningful data properties (dimensionality reduction). When...

As datasets continue to grow in size and complexity, analyzing every variable can quickly become challenging. Many datasets contain hundreds or even thousands of features, some often provide similar information. Principal Component Analysis, also known as PCA, is an...

When working with large amounts of data, one of the hardest challenges is not knowing where to start. Data can contain thousands to millions of records, and without clear labels, it can be difficult to understand how those records relate to one another. K-Means...

When analyzing large datasets, some of the most valuable insights come from discovering relationships that might otherwise go unnoticed. This is where the Apriori algorithm comes in. Supported within Oracle Machine Learning, it is an unsupervised algorithm designed to...

When people think about machine learning, they often focus on advanced algorithms such as neural networks or random forests. However, some of the most useful predictive models are built upon well established statistical techniques. One example is the Generalized...

Support Vector Machine (SVM) is a supervised machine learning algorithm that can be applied to both classification and regression problems. Its purpose is to detect patterns within data and make predictions by identifying the most effective boundary between different...

Exponential Smoothing is a forecasting algorithm used to analyze time based data and predict future values. It’s particularly useful when data is collected over regular intervals, such as daily sales figures, monthly revenue, or weekly demand. The algorithm works by...

Naive Bayes is a supervised machine learning algorithm consisting of three main types (Gaussian, Multinomial & Bernoulli). This probability-based algorithm is mainly used for classification tasks by using the relationships between different features to predict the...

Random Forest is a supervised machine learning algorithm used for both classification and regression tasks. It is based on the idea of combining multiple decision trees to produce a stronger and more reliable model. Instead of relying on a single tree, Random Forest...

Minimum Description Length, often shortened to MDL, is a principle used in machine learning and data analysis to help identify models that best explain a dataset while avoiding unnecessary complexity. The main idea behind MDL is that the best model is one that...

Neural networks are a machine learning technique inspired by the way the human brain processes information. They are designed to recognize patterns, learn from data, and make predictions by analyzing relationships between inputs and outputs. Neural networks are...

Decision trees are a supervised machine learning technique used for both classification and regression tasks. They work by breaking down a problem into a series of simple decisions, creating a tree like structure of rules that lead to a final outcome. Each step in the...

XGBoost, short for Extreme Gradient Boosting, is a powerful machine learning algorithm used mainly for supervised learning tasks such as classification and regression. It is based on the idea of combining multiple simple models, typically decision trees, to create a...

Ranking is a machine learning technique used to order items based on their relevance, importance, or likelihood of a particular outcome. Rather than simply predicting a value or assigning a category, ranking focuses on prioritizing results so that the most useful or...