Advanced Search
Search Results
210 total results found
Bayesian Networks
Bayesian Networks are data structure that represents the dependencies among random categorical variables: Directed graph Each node represents a random variable A direct edge from $X$ to $Y$ means $X$ is a parent of $Y$ Each node $X$ has probability distributi...
Uncertainty over time
We consider the Markov assumption: The current state $X_t$ depends on only a finite fixed number of previous states. A Markov chain is a sequence of random variablesfor which the distribution of each variable follows the Markov assumption. We can define a tran...
Example problems on uncertainty
Some example problems that could be asked in the exam: Given some knowledge about a problem, represent such knowledge as a Bayesian network Given a Bayesian network, specify what formula you need to compute a given probability Given a Bayesian network, comput...
linear models for regression
Questions Describe and compare Ridge regression and LASSO algorithms to solve linear regression problems. (8) These two regularization techniques differ from ordinary least squares because they introduce a model complexity penalization term in the loss functio...
linear models for classification
Questions Describe the Perceptron model and how it is trained. (1) The Perceptron algorithm is an example of online linear discriminant model. It corresponds to a two classes model: $$ y(x) = f(w^T \phi(x)) \text{, where } f(a) = \begin{cases} +1, a \ge 0 \ -1...
bias-variance, model selection and model ensembles
Questions Describe the PCA technique and what it is used for. (3) Principal Component Analysis (PCA) is a linear dimensionality reduction technique. The idea is to project the data onto the input subspace which accounts for the most variance. The steps of the ...
learning theory
Questions Explain what is the VC dimension of a hypothesis space and what it is used for. (6) As needed background we need to define: A dichotomy of a set $S$ is a partition of $S$ into two disjoint subsets A set of instances $S$ is shattered by the hypothesi...
kernel methods
CS480/680 Lecture 11: Kernel Methods Questions from past exams Give the definition of valid kernel and describe how valid kernels can be built. Provide an example of a methods that uses kernels and specify the advantages of using them in this specific meth...
list of questions
List of question and appearances as 2023-06-06: Question Appearances Topic Describe and compare Ridge regression and LASSO algorithms to solve linear regression problems. 9 02_Linear_Regression Describe Support Vector Machines (SVMs) for supervised cl...
Markov Decision Processes
Questions Describe the properties of the Bellman operators. (1) ... Describe the policy iteration technique for control problems on Markov Decision Processes. (3) Describe the value iteration algorithm and its properties. Does the algorithm always return the o...
reinforcement learning
Questions Describe and compare Q-learning and SARSA. (4) Describe the difference between on-policy and off-policy reinforcement learning techniques. Make an example of an on-policy algorithm and an example of an off-policy algorithm. (2) On-policy vs Off-polic...