Machine learning key terminologies
1. Basic Concepts
🔹 Machine Learning
A branch of Artificial Intelligence that enables systems to learn patterns from data and make
decisions without being explicitly programmed.
🔹 Dataset
A collection of data used for training and testing models.
Training Set – Data used to train the model.
Test Set – Data used to evaluate performance.
Validation Set – Data used to tune model parameters.
🔹 Features
Input variables (independent variables).
Example: In house price prediction → size, location, number of rooms.
🔹 Label (Target)
Output variable the model predicts.
Example: House price.
2. Types of Machine Learning
🔹 Supervised Learning
Model learns from labeled data.
Classification
Regression
🔹 Unsupervised Learning
Model finds patterns in unlabeled data.
Clustering
Dimensionality reduction
🔹 Reinforcement Learning
Agent learns by interacting with environment using rewards and penalties.
3. Core Algorithms
🔹 Linear Regression
Used for predicting continuous values.
🔹 Logistic Regression
Used for classification (especially binary classification).
🔹 Decision Tree
Tree-like structure used for classification and regression.
🔹 Random Forest
Ensemble of decision trees.
🔹 Support Vector Machine (SVM)
Finds optimal boundary between classes.
🔹 K-Nearest Neighbors (KNN)
Classifies based on closest data points.
🔹 Neural Networks
Inspired by the human brain; foundation of deep learning.
4. Model Training Terms
🔹 Model
Mathematical function that maps inputs (features) to outputs (labels).
🔹 Parameters
Internal variables learned during training (e.g., weights in neural networks).
🔹 Hyperparameters
Settings defined before training (e.g., learning rate, number of layers).
🔹 Loss Function
Measures how wrong the model’s predictions are.
🔹 Cost Function
Average loss across the dataset.
🔹 Optimization
Process of minimizing loss (e.g., Gradient Descent).
🔹 Learning Rate
Step size used during optimization.
5. Performance Metrics
🔹 Accuracy
Correct predictions / Total predictions.
🔹 Precision
Correct positive predictions / Total predicted positives.
🔹 Recall (Sensitivity)
Correct positive predictions / Total actual positives.
🔹 F1 Score
Harmonic mean of Precision and Recall.
🔹 Confusion Matrix
Table showing prediction results vs actual values.
🔹 Mean Squared Error (MSE)
Used for regression evaluation.
6. Overfitting & Underfitting
🔹 Overfitting
Model memorizes training data but performs poorly on new data.
🔹 Underfitting
Model too simple to capture patterns in data.
🔹 Regularization
Technique to prevent overfitting (L1, L2).
7. Advanced Concepts
🔹 Bias
Error due to overly simple assumptions.
🔹 Variance
Error due to model sensitivity to small fluctuations.
🔹 Bias-Variance Tradeoff
Balance between model complexity and generalization.
🔹 Cross-Validation
Technique to evaluate model stability.
🔹 Feature Engineering
Process of creating useful input features.
🔹 Dimensionality Reduction
Reducing number of features (e.g., PCA).
8. Deep Learning Specific Terms
🔹 Epoch
One full pass through the training dataset.
🔹 Batch Size
Number of samples processed before updating model.
🔹 Activation Function
Function applied to neural network output (ReLU, Sigmoid, Softmax).
🔹 Backpropagation
Algorithm for updating neural network weights.