SUPERVISED LEARNING
Definition
Category of machine learning where algorithms learn to make predictions or classify data based on labeled training dataset.
Characteristics
Predictive Modeling
Labelled data
Clear Objectives
Human Intervention
Categories
CLASSIFICATION
REGRESSION
Binary Classification
Types
Definition
A method where the model tries to predict the correct label of a given input data.
Spam/ Not Spam emails
Matched / Not Matched Face Recognition
Churn/ Stay for a service
Positive/ Negative Sentiments
Multi-class Classification
Sports/ Technology/ Politics/ Entertainment in News articles
Positive/ Negative/ Neutral in feedback analysis
Defective/ Good/ Needs Repair in fault detection in manufacturing
Cat/ Dog/ Bird/ Elephant in image detection
Definition
A technique used to capture the relationships between independent and dependent variables, with the main purpose of predicting an outcome
Multiple Linear Regression
Logistic Regression
Types
Simple Linear Regression
Predicting if a credit card transaction is fraudulent or not fraudulent
Salary and Experience
Analyzing diseases based on variables like age, gender, weight, diet, exercise, and medication.
Algorithms
Support Vector Machines
Decision Tree
K - Nearest Neighbor
Simple to understand and implement.
Predictions are based on the majority class (classification) or average (regression) of the K nearest neighbors.
Sensitive to the choice of distance metric (e.g., Euclidean, Manhattan).
Classification and regression can be performed using KNN.
Applications of Supervised Learning
Recommendation System
Medical Diagnosis
Face Detection
Suitable for small to medium-sized datasets
Voice Recognition
Handwriting Recognition
Prone to overfitting with noisy data if not regularized properly
Objective of SVM is to find optimal hyperplane
Constructed by finding a hyperplane that maximizes the margin between different classes
Used for modeling decisions or rules in a tree-like structure.
Graphical representation for getting all the possible solutions to a problem/decision based on given conditions.
A decision tree simply asks a question, and based on the answer (Yes/No), it
further split the tree into subtrees.
Image-Based Diagnostics
Cancer Detection and Classification
Disease Detection
Archiving Handwritten Forms
Transcribing Medical Records
Handwriting-to-Text Apps
E-commerce Product Recommendations
Social Media Feeds
Movie and TV Show Recommendations
Music Recommendations
Voice Assistants
Voice Typing
Language Translation
Voice Biometrics
Face Detection in Cameras and Smartphones
Social Media Photo Tagging
Identity Verification
Agriculture
Crop Disease Detection
Crop Yield Prediction
Irrigation Optimization