A machine learningtask is a type of prediction or inference that's based on both:

• The problem or question
• The available data

For example, the classification task assigns data to categories, and the clustering task groups data according to similarity.

Machine learning tasks rely on patterns in the data rather than being explicitly programmed.

This article describes the different machine learning tasks that are available in ML.NET and some common use cases.

Once you've decided which task works for your scenario, then you need to choose the best algorithm to train your model. The available algorithms are listed in the section for each task.

Classification tasks

• Binary classification
• Multiclass classification
• Image classification

Binary classification

Binary classification is asupervised machine learning task that's used to predict which ofexactly two classes (categories) an instance of data belongs to. The input of a classification algorithm is a set of labeled examples, where each label is an integer of either 0 or 1. The output of a binary classification algorithm is a classifier, which you can use to predict the class of new unlabeled instances. Examples of binary classification scenarios include:

• Understanding sentiment of Twitter comments as either positive or negative.
• Diagnosing whether a patient has a certain disease.
• Making a decision to mark an email as spam.
• Determining if a photo contains a particular item, such as a dog or fruit.

For more information, see theBinary classification article on Wikipedia.

Binary classification trainers

You can train a binary classification model using the following algorithms:

• AveragedPerceptronTrainer
• SdcaLogisticRegressionBinaryTrainer
• SdcaNonCalibratedBinaryTrainer
• SymbolicSgdLogisticRegressionBinaryTrainer
• LbfgsLogisticRegressionBinaryTrainer
• LightGbmBinaryTrainer
• FastTreeBinaryTrainer
• FastForestBinaryTrainer
• GamBinaryTrainer
• FieldAwareFactorizationMachineTrainer
• PriorTrainer
• LinearSvmTrainer

Binary classification inputs and outputs

For best results with binary classification, the training data should be balanced (that is, equal numbers of positive and negative training data). Missing values should be handled before training.

The input label column data must beBoolean.The input features column data must be a fixed-size vector ofSingle.

These trainers output the following columns:

Multiclass classification

Multiclass classification is asupervised machine learning task that's used to classify an instance of data into one ofthree or more classes (categories). The input of a classification algorithm is a set of labeled examples. Each label normally starts as text. It's then run through the TermTransform, which converts it to the Key (numeric) type. The output of a classification algorithm is a classifier, which you can use to predict the class of new unlabeled instances. Examples of multiclass classification scenarios include:

• Categorizing flights as "early", "on time", or "late".
• Understanding movie reviews as "positive", "neutral", or "negative".
• Categorizing hotel reviews as "location", "price", or "cleanliness", for example.

For more information, see theMulticlass classification article on Wikipedia.

Multiclass classification trainers

You can train a multiclass classification model using the following training algorithms:

• TextClassificationTrainer
• LightGbmMulticlassTrainer
• SdcaMaximumEntropyMulticlassTrainer
• SdcaNonCalibratedMulticlassTrainer
• LbfgsMaximumEntropyMulticlassTrainer
• NaiveBayesMulticlassTrainer
• OneVersusAllTrainer
• PairwiseCouplingTrainer

Multiclass classification inputs and outputs

The input label column data must bekey type.The feature column must be a fixed size vector ofSingle.

This trainer outputs the following:

Text classification

Text classification is a subcategory of multiclass classification that deals specifically with raw text. Text poses interesting challenges because you have to account for the context and semantics in which the text occurs. As such, it can be difficult to encode meaning and context.

Deep learning models have emerged as a promising technique to solve natural language problems. More specifically, a type of neural network known as atransformer has become the predominant way of solving natural language problems like text classification, translation, summarization, and question answering. Some popular transformer architectures for natural language tasks include:

• Bidirectional encoder representations from transformers (BERT)
• Robustly optimized BERT pretraining approach (RoBERTa)
• Generative pretrained transformer (GPT)

The ML.NET text classification API is powered byTorchSharp. TorchSharp is a .NET library that provides access to the library that powers PyTorch. TorchSharp contains the building blocks for training neural networks from scratch in .NET. ML.NET abstracts some of the complexity of TorchSharp to the scenario level. It uses a pretrained version of theNAS-BERT model and fine tunes it with your data.

For a text classification example, seeGet started with the text classification API.

Image classification

Image classification is asupervised machine learning task that's used to predict the class (category) of an image. The input is a set of labeled examples. Each label normally starts as text. It's then run through the TermTransform, which converts it to the Key (numeric) type. The output of the image classification algorithm is a classifier, which you can use to predict the class of new images. The image classification task is a type of multiclass classification. Examples of image classification scenarios include:

• Determining the breed of a dog as a "Siberian Husky", "Golden Retriever", "Poodle", etc.
• Determining if a manufacturing product is defective or not.
• Determining what types of flowers as "Rose", "Sunflower", etc.

Image classification trainers

You can train an image classification model using the following training algorithms:

• ImageClassificationTrainer

Image classification inputs and outputs

The input label column data must bekey type.The feature column must be a variable-sized vector ofByte.

This trainer outputs the following columns:

Regression

Regression is asupervised machine learning task that's used to predict the value of the label from a set of related features. The label can be of any real value and isn't from a finite set of values as in classification tasks. Regression algorithms model the dependency of the label on its related features to determine how the label will change as the values of the features are varied. The input of a regression algorithm is a set of examples with labels of known values. The output of a regression algorithm is a function, which you can use to predict the label value for any new set of input features. Examples of regression scenarios include:

• Predicting house prices based on house attributes such as number of bedrooms, location, or size.
• Predicting future stock prices based on historical data and current market trends.
• Predicting sales of a product based on advertising budgets.
• Finding related articles in a publication (sentence similarity).

Regression trainers

You can train a regression model using the following algorithms:

• SentenceSimilarityTrainer
• LbfgsPoissonRegressionTrainer
• LightGbmRegressionTrainer
• SdcaRegressionTrainer
• OlsTrainer
• OnlineGradientDescentTrainer
• FastTreeRegressionTrainer
• FastTreeTweedieTrainer
• FastForestRegressionTrainer
• GamRegressionTrainer

Regression inputs and outputs

The input label column data must beSingle.

The trainers for this task output the following:

Clustering

Clustering is anunsupervised machine learning task that's used to group instances of data into clusters that contain similar characteristics. Clustering can also be used to identify relationships in a dataset that you might not logically derive by browsing or simple observation. The inputs and outputs of a clustering algorithm depend on the methodology chosen. You can take a distribution, centroid, connectivity, or density-based approach. ML.NET currently supports a centroid-based approach using K-Means clustering. Examples of clustering scenarios include:

• Understanding segments of hotel guests based on habits and characteristics of hotel choices.
• Identifying customer segments and demographics to help build targeted advertising campaigns.
• Categorizing inventory based on manufacturing metrics.

Clustering trainer

You can train a clustering model using the following algorithm:

• KMeansTrainer

Clustering inputs and outputs

The input features data must beSingle. No labels are needed.

This trainer outputs the following:

Anomaly detection

The anomaly detection task creates an anomaly detection model by using principal component analysis (PCA). PCA-based anomaly detection helps you build a model in scenarios where it's easy to obtain training data from one class, such as valid transactions, but difficult to obtain sufficient samples of the targeted anomalies.

An established technique in machine learning, PCA is frequently used in exploratory data analysis because it reveals the inner structure of the data and explains the variance in the data. PCA works by analyzing data that contains multiple variables. It looks for correlations among the variables and determines the combination of values that best captures differences in outcomes. These combined feature values are used to create a more compact feature space called the principal components.

Anomaly detection encompasses many important tasks in machine learning:

• Identifying transactions that are potentially fraudulent.
• Learning patterns that indicate that a network intrusion has occurred.
• Finding abnormal clusters of patients.
• Checking values entered into a system.

Because anomalies are rare events by definition, it can be difficult to collect a representative sample of data to use for modeling. The algorithms included in this category have been especially designed to address the core challenges of building and training models by using imbalanced data sets.

Anomaly detection trainer

You can train an anomaly detection model using the following algorithm:

• RandomizedPcaTrainer

Anomaly detection inputs and outputs

The input features must be a fixed-sized vector ofSingle.

This trainer outputs the following:

Ranking

A ranking task constructs a ranker from a set of labeled examples. This example set consists of instance groups that can be scored with a given criteria. The ranking labels are { 0, 1, 2, 3, 4 } for each instance. The ranker is trained to rank new instance groups with unknown scores for each instance. ML.NET ranking learners aremachine-learned ranking based.

Ranking training algorithms

You can train a ranking model with the following algorithms:

• LightGbmRankingTrainer
• FastTreeRankingTrainer

Ranking input and outputs

The input label data type must bekeytype orSingle. The value of the label determines relevance, wherehigher values indicate higher relevance. If the label is akey type, then the key index is therelevance value, where the smallest index is the least relevant. If the label is aSingle, larger values indicate higher relevance.

The feature data must be a fixed size vector ofSingle and input row groupcolumn must bekey type.

This trainer outputs the following:

Recommendation

A recommendation task enables producing a list of recommended products or services. ML.NET usesMatrix factorization (MF), acollaborative filtering algorithm for recommendations when you have historical product rating data in your catalog. For example, you have historical movie rating data for your users and want to recommend other movies they're likely to watch next.

Recommendation training algorithms

You can train a recommendation model with the following algorithm:

• MatrixFactorizationTrainer

Forecasting

The forecasting task use past time-series data to make predictions about future behavior. Scenarios applicable to forecasting include weather forecasting, seasonal sales predictions, and predictive maintenance.

Forecasting trainers

You can train a forecasting model with the following algorithm:

ForecastBySsa

Object detection

Object detection is asupervised machine learning task that's used to predict the class (category) of an image but also gives a bounding box to where that category is within the image. Instead of classifying a single object in an image, object detection can detect multiple objects within an image. Examples of object detection include:

• Detecting cars, signs, or people on images of a road.
• Detecting defects on images of products.
• Detecting areas of concern on X-Ray images.

Object-detection model training is currently only available inModel Builder using Azure Machine Learning.