Logistic Regression is commonly used to solve various classification problems such as spam detection.

It is one of the most simple machine learning algorithms for two-class classification. While linear regression gives a continuous output (housing prices), logistic regression provides a discrete output (whether a patient has cancer or not).

Aside from its easy implementation, it is also efficient and does not require high computational power. However, logistic regression does not do well with a large number of features and also has a tendency to overfit.

K Nearest Neighbors is a supervised machine learning algorithm that is commonly used to solve classification problems.

For this algorithm, a key assumption made is that similar things exist in close proximity. By calculating the distances between various data points, the model will attempt to predict the class of a new data point based on its proximity to other points.

The advantage to such an approach is that it is simple to implement with little need for us to tune any parameters. However, the algorithm gets significantly slower with increasing number of features used in the model.

XGBoost(eXtreme Gradient Boosting) is a decision-tree-based ensemble machine learning algorithm that uses a gradient boosting framework.

Designed for speed and performance, it has become a widely used and really popular tool among Kaggle competitors and Data Scientists in industry. XGBoost uses an ensemble method that seeks to create a strong classifier (model) based on “weak” classifiers. Models are added on top of each other iteratively and the errors of the previous model are then corrected by the next predictor.

However, as a tradeoff to its laudable efficiency and performance, it is significantly harder to tune than other models due to the many hyperparameters and is susceptibie to overfitting.