Roll Number
12i190005
Category
TA
Topics for PhD Qualifiers
Compulsory Subject: (i) Optimisation Techniques, (ii) Stochastic Models
Elective 1: Data Mining & Machine Learning
Elective 2: Markov Decison Processes
Elective 1: Data Mining & Machine Learning
Elective 2: Markov Decison Processes
Elective1
Data mining & Machine learning
a) Basic Statistics( testing and estimation procedures), Bayesian statistics, Regression & Logistic Regression.
b) Biased Regression Techniques : Ridge & LASSO , Penalized Regression, Splines.
c) Latent Linear Models : Factor Analysis, Principal component analysis, Independent component analysis.
d) Sparse Linear Models: Bayesian Variable selection, L1 regularization basics and algorithms.
e) Kernels: Kernel functions, Support vector machines.
a) Basic Statistics( testing and estimation procedures), Bayesian statistics, Regression & Logistic Regression.
b) Biased Regression Techniques : Ridge & LASSO , Penalized Regression, Splines.
c) Latent Linear Models : Factor Analysis, Principal component analysis, Independent component analysis.
d) Sparse Linear Models: Bayesian Variable selection, L1 regularization basics and algorithms.
e) Kernels: Kernel functions, Support vector machines.
Elective2
Markov Decsion Processes
a) Finite Horizon & Infinite Horizon Markov Decision Processes : Optimality criteria & examples.
b) Discounted MDP : Optimality Equations, Value iteration, Policy iteration, Modified Policy iteration, Linear programming as MDP.
c) Average Reward criterion (Unichain Models) : Optimality Equations, Value iteration, policy iteration, modified policy iteration and Linear programmming.
d) Continuous time Models : Discounted models, average reward models,continuous time Markov decision processes, Queing admission control .
a) Finite Horizon & Infinite Horizon Markov Decision Processes : Optimality criteria & examples.
b) Discounted MDP : Optimality Equations, Value iteration, Policy iteration, Modified Policy iteration, Linear programming as MDP.
c) Average Reward criterion (Unichain Models) : Optimality Equations, Value iteration, policy iteration, modified policy iteration and Linear programmming.
d) Continuous time Models : Discounted models, average reward models,continuous time Markov decision processes, Queing admission control .
PhD. Supervisor (if decided)
Prof. N. Hemachandra