Machine Learning Midterm

Types of Machine Learning

• Supervised Learning uses labeled data to train models, aiming to predict output from labeled input data.
• Primary goal of Supervised Learning is to predict output from labeled input data.

Classification Algorithms

• Logistic Regression is a common algorithm used in Classification.

Clustering Algorithms

• K-Means Clustering is used to segment customers into groups.

Model Evaluation

• Cross-Validation ensures the model generalizes well to unseen data.

Neural Networks

• An Epoch is a single forward and backward pass of all the training examples.

Classification Model Metrics

• Mean Squared Error is not a metric used in Classification models.

Overfitting

• Overfitting occurs when a model performs well on training data but poorly on test data.

Activation Functions

• Activation Functions introduce non-linearity into the model.

Feature Selection Methods

• Principal Component Analysis (PCA) is an example of a Feature Selection Method.

Regression Algorithms

• Linear Regression is a common algorithm used in Regression.

Performance Evaluation

• Confusion Matrix is used to evaluate the performance of a classification algorithm.

Sequential Decision Making

• Reinforcement Learning is used for Sequential Decision Making.

Data Preprocessing

• Normalization is used to scale data to a standard range.

Deep Learning Models

• Deep Learning Models automatically learn feature representations.
• Convolutional Neural Networks (CNNs) are commonly used for Image recognition.

Hyperparameters

• Hyperparameters are set before the learning process begins.

Ensemble Learning

• Ensemble Learning uses multiple models to achieve better performance.

Unsupervised Learning

• K-Means Clustering is an example of Unsupervised Learning.

Curse of Dimensionality

• Curse of Dimensionality refers to the problem of overfitting in high-dimensional spaces.

Reinforcement Learning

• Policy is the strategy that the agent employs to determine its actions.

Imbalanced Classification

• Precision-Recall Curve is an appropriate metric for Imbalanced Classification.

Backpropagation Algorithm

• Backpropagation Algorithm updates the weights by minimizing the loss function.

Classification Errors

• False Positive occurs when a negative class is incorrectly identified as positive.

Handling Missing Data

• Imputation is a technique used to handle missing data.

k-Nearest Neighbors (k-NN) Algorithm

• k-NN Algorithm uses a distance metric to classify new points.

Bootstrapping

• Bootstrapping is a resampling technique used to estimate statistics on a population.

Preventing Overfitting

• Regularization techniques are used to prevent overfitting.

Dropout

• Dropout is a regularization method used to prevent overfitting in Neural Networks.

Principal Component Analysis (PCA)

• PCA transforms data into a lower-dimensional space.