What is an ML Model? - AWS AI Practitioner AIF-C01

Defining Models in General
📌 A model is generally an information representation of an object, person, or system.
🧱 Models can be concrete, meaning they have a physical form, such as a vehicle design or a photograph.
📊 Abstract models are expressed as behavioral patterns, like mathematical equations, computer code, or written words.

Machine Learning (ML) Model Definition and Process
⚙️ An ML model is a function that takes input data and applies a machine learning algorithm to generate a prediction.
🎓 The term training model refers to the process where learning occurs to make correct predictions, involving training data (often labeled) and a learning algorithm.
🛠️ Hyper-tuning is a continuous process of tweaking the model during training to optimize its performance.
🚀 Once tuned and deployed, the ML model becomes the trained model, capable of producing predictions, which is the interaction known as inference.

Inference in Machine Learning
🗣️ Inference is the interaction with a deployed machine learning model where data (often unlabeled) is provided to solicit a prediction.
🎯 The main goal of the deployed ML model is to produce predictions based on the data it receives during inference.

Key Points & Insights
➡️ A model serves as an information representation that can be either concrete (physical) or abstract (behavioral/code).
➡️ The training model undergoes optimization through hyper-tuning using labeled data and a learning algorithm.
➡️ Inference is the operational phase where a deployed, trained ML model uses input data to make its final predictions.

📸 Video summarized with SummaryTube.com on Feb 03, 2026, 02:56 UTC