Week 0 [Optional]

What is a confusion matrix?

A confusion matrix is a table used to evaluate the performance of a classification model. It shows the counts of true positives, true negatives, false positives, and false negatives, providing insights into accuracy, precision, recall, and F1-score.

What is a decision boundary in machine learning?

A decision boundary is the surface that separates different classes in a classification problem. It defines the regions where the model classifies data points as belonging to one class or another.

What is a hyperparameter?

A hyperparameter is a parameter whose value is set before the learning process begins. It controls the behavior of the training algorithm, such as learning rate, number of layers, or batch size. These are different from model parameters, which are learned from data.

What is a loss function?

A loss function measures how well a machine learning model's predictions match the actual target values. It computes the error between predicted and true values and guides the optimization process in reducing this error.

What is a neural network?

A neural network is a set of algorithms, designed to recognize patterns. It consists of interconnected layers of nodes (neurons) that process input data, learn weights and biases, and produce output through activation functions.

What is a validation set?

A validation set is a subset of the training data used to tune hyperparameters and monitor the model's performance during training. It helps prevent overfitting by ensuring the model generalizes well before testing it on a separate test set.

What is the activation function in neural networks?

An activation function is used in neural networks to introduce non-linearity into the model, allowing it to learn complex patterns. Common activation functions include ReLU (Rectified Linear Unit), sigmoid, and tanh.

What is backpropagation?

Backpropagation is the process used in training neural networks where the gradient of the loss function is computed for each weight by the chain rule, working backward from the output to the input. This helps update weights to minimize the loss.

What is generalization in machine learning?

Generalization refers to a model's ability to perform well on new, unseen data. A model that generalizes well captures the underlying patterns in the data rather than memorizing it, avoiding overfitting.

What is gradient descent?

Gradient descent is an optimization algorithm used to minimize the loss function in machine learning models. It works by updating model parameters (weights) iteratively in the direction of the negative gradient of the loss function with respect to the parameters.

What is regularization?

Regularization is a technique used to prevent overfitting by adding a penalty term to the loss function. Common types include L1 regularization (Lasso), which encourages sparsity, and L2 regularization (Ridge), which discourages large weight values.

What is reinforcement learning?

Reinforcement learning is a type of machine learning where an agent learns to make decisions by performing actions in an environment to maximize cumulative rewards. The agent explores different actions and receives feedback based on its performance.

What is stochastic gradient descent (SGD)?

Stochastic gradient descent (SGD) is a variant of gradient descent where the model parameters are updated using a single (or a small batch of) training example(s) at each iteration. This introduces noise but can lead to faster convergence compared to batch gradient descent.

What is supervised learning?

Supervised learning is a type of machine learning where the model is trained on labeled data. The goal is to learn a mapping from input features (X) to target labels (Y), so it can predict labels for new, unseen data.

What is the bias-variance tradeoff?

The bias-variance tradeoff refers to the balance between two sources of error in machine learning models: • Bias: Error due to overly simplistic models that underfit the data. • Variance: Error due to overly complex models that overfit the data. A good model minimizes both bias and variance to achieve good generalization.

What is a learning rate?

The learning rate is a hyperparameter that controls the step size during the update of model weights in gradient descent. A higher learning rate leads to faster updates but may cause overshooting, while a lower learning rate leads to slower updates but can converge more precisely.

What is unsupervised learning?

Unsupervised learning involves training a model on data that doesn't have labeled responses. The goal is to find hidden patterns, structures, or relationships in the data (e.g., clustering or dimensionality reduction).

What are the key differences between AI, ML, and Deep Learning?

• AI (Artificial Intelligence): Broad field focused on creating systems that can perform tasks typically requiring human intelligence. • ML (Machine Learning): A subset of AI that focuses on systems that can learn from data and improve without being explicitly programmed. • Deep Learning: A subset of ML that uses neural networks with multiple layers to model complex patterns in data.

What is the difference between classification and regression?

• Classification: Predicts discrete labels or categories (e.g., yes/no, cat/dog). • Regression: Predicts continuous values (e.g., predicting house prices, stock market prices).

What is overfitting and underfitting?

• Overfitting: When a model learns not only the underlying patterns but also the noise in the training data, leading to poor performance on new data. • Underfitting: When a model is too simple to capture the underlying patterns in the data, resulting in poor performance on both training and new data.