COMP SCI 540
Artificial Intelligence
University of Wisconsin - Madison
University of Wisconsin - Madison
Understand and be able to apply the foundational tools in Machine Learning and Artificial Intelligence: Linear algebra, Probability, Logic, and elements of Statistics.
Understand core techniques in Natural Language Processing (NLP), including bag-of-words, tf-idf, n-Gram Models, and Smoothing.
Understand the basics of Machine Learning. Identify and summarize important features in supervised learning and unsupervised learning.
Distinguish between regression and classification, and understand basic algorithms: Linear Regression, k-Nearest Neighbors, and Naive Bayes.
Understand the basics of Neural Networks: Network Architecture, Training, Backpropagation, Stochastic Gradient Descent.
Learn aspects of Deep Learning, including network architectures, convolution, training techniques.
Understand the fundamentals of Game Theory.
Understand how to formulate and solve several types of Search problems.
Understand basic elements of Reinforcement Learning.
Consider how Artificial Intelligence and Machine Learning problems are applied in Real - World settings and the Ethics of Artificial Intelligence.
EXAM 1 REVIEW MATERIAL
Topics:
Week 1: Basic probability - joint probability, conditional probability, and Bayesian rule
Week 1: Statistics, Linear Algebra, and PCA - empirical estimation, linear algebra review, dimensionality reduction, and PCA
Week 2: Logic and Natural Language Processing - logic review, knowledge bases, NLP goals, NLP n-gram models, and perplexity
Week 3: Machine Learning Overview - machine learning overview, supervised learning overview, and unsupervised learning overview
Week 3/4: Unsupervised Learning - k-means clustering, hierarchical clustering, and t-SNE
Week 4: Linear Models and Regression - kernel density estimation, supervised learning model training, and measuring training loss with linear/logistic regression
Week 5: Classifications - K-NN algorithm and maximum likelihood estimation
Week 5: Perceptron - Naive Bayes, perceptrons, and linear perceptrons
Week 6: Neural Networks - activation functions, multilayer perceptrons, and gradient descent