Azure AI Data and AI Models Flashcards
| Acronym | Definition |
|---|---|
| What is Data Preprocessing | The process of cleaning transforming and preparing raw data for use in AI models |
| What are the main steps in Data Preprocessing | Cleaning transformation splitting into training and test sets |
| What is Data Cleaning | Removing or correcting inaccurate incomplete or irrelevant data |
| What is Feature Transformation | Altering data to improve its suitability for machine learning often including scaling or encoding |
| Why is Splitting Data important | To separate data into training and testing sets ensuring model evaluation is unbiased |
| What is Training in AI | The process of feeding data into a machine learning algorithm to create a model |
| What is Inference in AI | Using a trained model to make predictions or decisions based on new data |
| What are common evaluation metrics for Classification models | Accuracy precision recall and F1 score |
| What are common evaluation metrics for Regression models | Mean Absolute Error Mean Squared Error and R squared |
| What is Model Overfitting | When a model learns the training data too well and performs poorly on unseen data |
| What is Model Underfitting | When a model is too simple and fails to capture patterns in the training data |
| What is Hyperparameter Tuning | The process of optimizing settings for an AI model to improve its performance |
| What is Batch Deployment | Deploying an AI model to process data in batches rather than real time |
| What is Real-time Deployment | Deploying an AI model to provide immediate predictions or decisions as data is received |
| Which Azure service supports model training and deployment | Azure Machine Learning |
| What is the purpose of Cross-validation | To ensure the model generalizes well by testing it on different subsets of the training data |
| What is a Pipeline in Azure Machine Learning | A series of steps to automate workflows for data preparation training and deployment |
| What is Responsible AI in Azure | Ensuring AI models are fair explainable and ethically deployed |
| What is the difference between Training and Inference | Training builds the model while inference uses the model to make predictions |
| What is Feature Engineering | The process of selecting creating or transforming variables to improve model performance |
Front
What is Real-time Deployment
Click the card to flip
Back
Deploying an AI model to provide immediate predictions or decisions as data is received
Front
What are common evaluation metrics for Regression models
Back
Mean Absolute Error Mean Squared Error and R squared
Front
What is a Pipeline in Azure Machine Learning
Back
A series of steps to automate workflows for data preparation training and deployment
Front
What is the purpose of Cross-validation
Back
To ensure the model generalizes well by testing it on different subsets of the training data
Front
What are common evaluation metrics for Classification models
Back
Accuracy precision recall and F1 score
Front
What is Hyperparameter Tuning
Back
The process of optimizing settings for an AI model to improve its performance
Front
What is Feature Transformation
Back
Altering data to improve its suitability for machine learning often including scaling or encoding
Front
What is Training in AI
Back
The process of feeding data into a machine learning algorithm to create a model
Front
What is Feature Engineering
Back
The process of selecting creating or transforming variables to improve model performance
Front
What are the main steps in Data Preprocessing
Back
Cleaning transformation splitting into training and test sets
Front
What is Data Cleaning
Back
Removing or correcting inaccurate incomplete or irrelevant data
Front
What is Data Preprocessing
Back
The process of cleaning transforming and preparing raw data for use in AI models
Front
What is Inference in AI
Back
Using a trained model to make predictions or decisions based on new data
Front
What is Model Overfitting
Back
When a model learns the training data too well and performs poorly on unseen data
Front
Why is Splitting Data important
Back
To separate data into training and testing sets ensuring model evaluation is unbiased
Front
What is Batch Deployment
Back
Deploying an AI model to process data in batches rather than real time
Front
What is the difference between Training and Inference
Back
Training builds the model while inference uses the model to make predictions
Front
What is Model Underfitting
Back
When a model is too simple and fails to capture patterns in the training data
Front
Which Azure service supports model training and deployment
Back
Azure Machine Learning
Front
What is Responsible AI in Azure
Back
Ensuring AI models are fair explainable and ethically deployed
1/20