Lecture Notes on Machine Learning Concepts

Flashcards covering key concepts and terminology related to machine learning, particularly focused on concepts introduced in the lecture.

Macro Averaging

An averaging method that gives equal weightage to all classes, regardless of the number of samples in each class.

Imbalanced Dataset

A dataset where one class significantly outnumbers the others, making it difficult for a model to learn the minority class.

FM Score

A metric that combines precision and recall to evaluate the balance between the two.

UpSampling

A technique where the minority class is randomly sampled with repetition to increase its size for balancing the dataset.

DownSampling

A technique where the majority class is reduced in size by randomly removing samples to balance the dataset.

Synthetic Data Augmentation

The process of generating new samples using techniques like rotation, noise addition, or cropping, often used for image datasets.

SMOTE

Synthetic Minority Over-sampling Technique, used to create synthetic instances of the minority class.

Addison

A method focusing on generating samples that are difficult to classify, used for oversampling the minority class.

Weighted Loss Function

A technique applied in neural networks to assign different penalities to misclassification of minority and majority classes.

Confusion Matrix

A table used to evaluate the performance of a machine learning model, illustrating the true positives, true negatives, false positives, and false negatives.

Logistic Regression

A classification algorithm used for binary classification problems, which finds the best-fitting model by using a logistic function.

K-Nearest Neighbors (KNN)

A classification algorithm that classifies data points based on the classes of their nearest neighbors.

Support Vector Machine (SVM)

A supervised machine learning algorithm that aims to find a hyperplane that best separates classes in a high-dimensional space.

Hyperplane

A decision boundary that separates different classes in a dataset, generalizing to multiple dimensions.

Kernel Trick

A method in SVM that transforms data into a higher-dimensional space to make it linearly separable.

Soft Margin

A classification margin that allows some misclassification to improve model robustness and generalization.

Slack Variable (Gamma)

A variable in SVM that allows for some misclassification error, contributing to the soft margin.

C Parameter

A hyperparameter in SVM that controls the trade-off between maximizing the margin and minimizing misclassification.