Machine Learning
Interview Questions

In supervised learning, the algorithm is trained on labeled data, where inputs are paired with corresponding outputs. Unsupervised learning involves training on unlabeled data, and the algorithm discovers patterns and relationships without predefined output labels.

The curse of dimensionality refers to the challenges that arise when working with high-dimensional data. As the number of features increases, the amount of data needed to generalize accurately also increases, leading to increased computational complexity and the risk of overfitting.

The bias-variance tradeoff involves finding the right balance between model simplicity (bias) and flexibility (variance). A model with high bias may underfit, while a model with high variance may overfit. The goal is to minimize both bias and variance for optimal model performance.

Cross-validation is a technique used to assess the performance of a model by dividing the dataset into multiple subsets, training the model on different subsets, and evaluating its performance on the remaining data. It helps in obtaining a more reliable estimate of the model's generalization performance.

Feature engineering involves selecting, transforming, or creating input variables to improve model performance. It's crucial as the quality of features directly impacts a model's ability to learn and generalize from the data.

Regularization is a technique used to prevent overfitting by adding a penalty term to the model's cost function. It discourages complex models, helping to improve generalization on unseen data.

Decision trees make decisions by splitting the data based on features. Their advantages include interpretability and ease of visualization. Disadvantages include susceptibility to overfitting, especially in deep trees.

Ensemble learning involves combining predictions from multiple models to improve overall performance. An example is a Random Forest, which aggregates predictions from multiple decision trees to enhance predictive accuracy and robustness.

Precision is the ratio of correctly predicted positive observations to the total predicted positives, while recall is the ratio of correctly predicted positive observations to the total actual positives. They are metrics used to evaluate the performance of classification models.

Overfitting occurs when a model learns the training data too well, capturing noise and making it perform poorly on new, unseen data. Underfitting happens when a model is too simple to capture the underlying patterns in the training data, leading to poor performance on both training and test sets.

Imbalanced datasets occur when the distribution of classes is not equal. Techniques for handling imbalanced data include resampling methods (oversampling minority class or undersampling majority class), using different evaluation metrics, and exploring ensemble methods. Addressing imbalanced datasets is crucial to avoid biased models that favor the majority class.

Regression predicts continuous numerical values, while classification predicts discrete labels or categories. For example, predicting house prices is a regression task, while classifying emails as spam or not spam is a classification task.

A confusion matrix is a table that shows the performance of a classification algorithm. It displays true positive, true negative, false positive, and false negative values. From the confusion matrix, metrics such as accuracy, precision, recall, and F1 score can be calculated to evaluate the model's performance.

SVM is a supervised learning algorithm used for classification and regression tasks. It works by finding the hyperplane that best separates different classes in the feature space. SVM is applied in image classification, handwriting recognition, and bioinformatics, among other domains.

Cross-entropy loss, or log loss, measures the performance of a classification model whose output is a probability value between 0 and 1. It is commonly used in classification problems as it penalizes models more severely for confidently incorrect predictions, encouraging well-calibrated probability estimates.