Inference "Inference" is applying a trained model to unlabeled samples to obtain the corresponding targets. In other words, "inference" is the process of making predictions using a model.
MACHINE LEARNING
-
Validation: Assessing Model Performance on Unseen Data
By
–
Validation Validation is the process that lets us know whether a model is any good. Usually, we run a set of (unseen) labeled samples through a model to ensure that it can predict the targets.
-
Training: Building Models from Labeled Data
By
–
Training Training is a process that builds a model. We take labeled samples during training and let the model gradually learn the relationships between features and the target.
-
Understanding Models: Features, Rules, and Target Predictions
By
–
Model A model defines the relationship between features and the target. You can think of a model as a set of rules that, given certain features, determine the corresponding target. For example, given the bedrooms, bathrooms, and square footage, we get the price.
-
Unlabeled Samples in Supervised Learning: Features Without Targets
By
–
Unlabeled sample Unlabeled samples contain features, but they don't contain the target: (x, ?) The goal of supervised learning is to build a model that predicts the target of unlabeled samples.
-
Labeled Samples: Training Data with Features and Targets
By
–
Labeled sample Labeled samples are used to train and validate a model. These are usually represented as (x, y), where "x" is a vector containing all the features, and "y" is the corresponding target.
-
Samples in Machine Learning: Labeled and Unlabeled Data
By
–
Samples (also known as "examples" or "instances") A sample is a particular instance of data. It could be "labeled" (when it specifies the target) or "unlabeled" (when it doesn't.)
-
Understanding Features in Machine Learning Prediction Models
By
–
Features (also referred to as "x" or "variables") These are the input variables to our problem. We use these features to predict the target. For example: • pixels of a picture
• number of bedrooms of a house
• square footage of a house -
ML Deployment and Productivity Stagnation Over the Past Decade
By
–
For all the talk about ML destroying jobs, productivity had been improving fast long before we used any ML, and over the past 10 years we went from "very little ML" to to "ML deployed everywhere" and productivity stayed stagnant