Download presentation
Presentation is loading. Please wait.
1
What is Artificial Intelligence?
1956
2
AI Terminology Supervised Learning: Partitioning data based on labels (input/output pairs) Unsupervised Learning: Partitioning data based on data structure (input only) Reinforcement Learning: Supervised or unsupervised data partitioning based on a reward feedback from environment (input and/or output with reward)
3
AI terminology SUPERVISED Learning
Decision Trees: model of decisions and their possible consequences Naïve Bays Classifier: probabilistic classifier with assumptions between features consequences Least Squares Regression: fitting straight line with minimum distance SUPERVISED Learning Logistic Regression: fitting binomial data using logistic distribution Support Vector Machine: linear hyperplane to separate groups Ensemble Method: weighted vote of prediction from set of classifiers
4
UNSUPERVISED Learning
AI terminology (cont.) Clustering Algorithms: grouping similar objects Principal Component Analysis (PCA): orthogonal transformation to convert a set of possibly correlated variables into linear set of uncorrelated variables UNSUPERVISED Learning Examples: Centroid-based algorithms Connectivity-based algorithms Density-based algorithms Probabilistic Dimensionality Reduction Neural networks / Deep Learning Independent Component Analysis: separating multivariate signals into statistically independent/non-Gaussian subcomponents
5
REINFORCEMENT Learning
AI terminology (cont.) SARSA: State, action, reward, state, action. Reward based Markov decision process policy Monte Carlo: learns from episodes of experience/policies Q-learning: finds optimal policy to maximize reward REINFORCEMENT Learning
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.