Machine Learning: Frequently Asked Questions

FAQs

Choosing the right model and metrics make all the difference in ML. If you need guidance for this, you’re in the right place.

Models & Metrics

What is supervised learning?

Supervised learning is a type of machine learning where the model is trained on labeled data — meaning each input has a known output. Examples include classification and regression tasks.

What is unsupervised learning?

Unsupervised learning involves training a model on data without labeled outcomes. It is used to find patterns or groupings in the data, such as with clustering or dimensionality reduction.

What is classification?

Classification is a supervised learning task where the goal is to assign inputs to one of several predefined categories. Examples include spam detection or emotion recognition.

What is regression?

Regression is a supervised learning task where the model predicts a continuous value, such as price, temperature, or probability.

What are common evaluation metrics?

Evaluation metrics help assess model performance. For classification: accuracy, precision, recall, and F1 score are common. For regression: MSE (mean squared error) and RMSE (root mean squared error) are often used.

What is accuracy?

Accuracy is the proportion of correct predictions over total predictions. It’s useful when class distribution is balanced.

What is precision?

Precision is the proportion of true positives among all predicted positives. It’s useful when false positives are costly.

What is recall?

Recall is the proportion of true positives captured out of all actual positives. It’s useful when false negatives are costly.

What is F1 score?

The F1 score is the harmonic mean of precision and recall. It balances both concerns when there’s a trade-off.

What is MSE?

MSE (Mean Squared Error) is the average of the squared differences between predicted and actual values. It penalizes larger errors more.

What is RMSE?

RMSE (Root Mean Squared Error) is the square root of MSE, making it more interpretable as it’s in the same unit as the target variable.

What is a train-test split?

Train-test splitting divides your dataset into a portion used to train the model and another used to test its performance on unseen data.

Why do GPUs help with big data or deep learning?

GPUs can perform many calculations in parallel, making them much faster than CPUs for training large models, especially in deep learning.

Why does model deployment matter?

Deployment makes your model accessible — turning it into a service that users, apps, or systems can interact with. Without deployment, your model can’t provide value in real-world applications.

What is a CI/CD pipeline?

CI/CD (Continuous Integration / Continuous Deployment) is a DevOps approach that automates testing, integration, and deployment. It ensures your models and code can be updated and delivered reliably and quickly.