Logistic Regression is mainly used to predict binary outcomes (e.g., yes/no, true/false). It models the probability that the dependent variable belongs to a particular category.

Gradient Boosting is a boosting method that builds decision trees sequentially. Each tree tries to correct the errors of the previous one by using gradients (direction of the steepest ascent) to minimize the loss function. This leads to a powerful model with improved accuracy.

Ensemble methods like Random Forest and Gradient Boosting combine the strengths of multiple models to achieve better accuracy and robustness. By leveraging a diverse set of models, they often outperform single models, especially on complex tasks, and reduce the risks of overfitting.

Considering interaction effects is essential to capture complex relationships between variables that might not be apparent when considering each variable separately.

Sentiment analysis in customer feedback systems often involves text classification techniques. Machine learning algorithms like SVM, Naïve Bayes, or deep learning models can categorize customer comments into positive, negative, or neutral sentiment.

Scaling features in KNN ensures that each feature contributes equally to the distance computation, preventing features with larger scales from dominating.

The Confusion Matrix is a table that describes the performance of a classification model by categorizing predictions into True Positives, False Positives, True Negatives, and False Negatives. It gives detailed insight into where the model is making mistakes.

A "testing" set consists of input variables and corresponding target values used to assess a machine learning model's performance on unseen data, allowing for a more robust evaluation.

The assumption of linearity ensures that the relationship between the independent and dependent variable is linear, which is fundamental to Simple Linear Regression.

Supervised Learning uses labeled data, where the output is known, to train the algorithm and make predictions.