- A machine learning algorithm is a method an AI system uses to predict output values from given input data.
- Classification and regression are the main supervised tasks; clustering, dimensionality reduction and anomaly detection are the main unsupervised ones.
- Each algorithm optimizes a specific loss function, and more data usually improves its performance.
- An algorithm is the process run on data; the model is what the algorithm learned.
- No single algorithm is best: the choice depends on the data, interpretability needs and available compute time.
What are Machine Learning Algorithms?
A machine learning algorithm is a method where the artificial intelligence system conducts a task of predicting output values from given input data. The two main tasks in supervised machine learning algorithms are classification and regression, while the main tasks in unsupervised machine learning are clustering, dimensionality reduction and anomaly detection.
What are Common Machine Learning Algorithms?
Below is a list of common machine learning algorithms:
- Linear Regression
- Logistic Regression
- Decision Tree
- Support Vector Machines
- Naive Bayes
- Nearest Neighbors
- K-Means Clustering
- Dimensionality Reduction
- Random Forest
- Gradient Boosting
- Neural Networks
- Ensembles of ML algorithms
What are Machine Learning Algorithms Used For?
Machine learning algorithms analyze data and create output values. The goal of each algorithm is to optimize a specific loss function and to improve its performance over time. More data usually improves the performance. The final state of the algorithm (the model) contains key insights about the data and forms the basis of the artificial intelligence.
Why are Machine Learning Algorithms Important?
Algorithms are core to machine learning solutions. Data scientists use algorithms as building blocks for more efficient problem-solving. A major benefit of machine learning processes is that they are time and cost-effective solutions that are executed with little or no human intervention.
Machine Learning Algorithms vs Other Technologies & Methodologies
Machine learning algorithms vs models
A machine learning algorithm is a process that is run on data to create a machine learning model. A model represents what was learned by the algorithm.
Machine learning algorithms vs artificial intelligence
Machine learning is a subset of artificial intelligence.
Machine learning algorithms vs deep learning algorithms
Machine learning is an application of artificial intelligence that includes algorithms that parse data, learn from that data, and then apply what they have learned onto new or unseen data to make predictions that inform decisions.
A traditional machine learning algorithm can be as simple as linear regression. By way of example, imagine you want to predict your income given your years of higher education. The first step is to define a function, e.g. income = y + x * years of education. Next, give the algorithm a set of training data. Then, let your algorithm draw the line, e.g. through an ordinary least squares (OLS) regression. Now, you can give the algorithm some test data, e.g. your personal years of higher education, and let it predict your income.
Deep learning is a subfield of machine learning that builds algorithms in layers to create an artificial neural network that can learn and make decisions on its own. Deep learning is a way to describe algorithms that analyze data with a logic structure similar to how a human mind would make conclusions. To achieve this, deep learning applications use an artificial neural network (ANN) which is a layered structure of algorithms.
Machine Learning Algorithms in H2O AI
Open-source H2O-3 implements many of these algorithms, including GLM, gradient boosting, distributed random forest, XGBoost, naive Bayes, stacked ensembles and AutoML for supervised learning, and k-means, PCA and isolation forest for unsupervised learning. H2O-3 algorithms documentation.