FREE
AKTU Optimization in Machine Learning (BAI071) Syllabus PDF – B.Tech 4th Year
Optimization in Machine Learning BAI071
All syllabus, notes, question bank & PYQ for this subject
Free
Download now
- English
- Semester 7
- 0 downloads
- 0 views
Official AKTU syllabus of Optimization in Machine Learning (BAI071) for B.Tech 4th Year – 3 credits, L-T-P 3-0-0. Read the unit-wise course content, course outcomes, evaluation pattern and recommended books online, or download the PDF.
Syllabus — Optimization in Machine Learning
Download PDFBAI071 · 3 credits · L-T-P 3-0-0
Course Outcomes
At the end of course , the student will be able to understand
| CO 1 | Understand the basics of the convex optimization. | K 2 |
| CO 2 | Understand the different Gradient-based methods. | K , K 2 3 |
| CO 3 | Can implement Newton’s method and L-BFGS solvers for convex optimization problems, | K , K 3 4 |
| CO 4 | Can identify the trade-offs inherent in using first-order vs. second-order solvers for optimization problems arising in machine learning. | K , K 2 3 |
| CO 5 | Demonstrate competence with probability theory/statistics needed to formulate and solve machine learning problems. | K , K 2 4 |
Course Content
L-T-P: 3-0-0
| Unit | Topic | Proposed Lecture |
|---|---|---|
| I | Basics of convex optimization Convex sets, convexity-preserving operations, examples of convex programs (linear programming (LP), second-order cone programming (SOCP), semidefinite programming (SDP)), convex relaxation, KKT conditions, duality | 09 |
| II | Gradient-based methods Gradient descent, subgradient, mirror descent, Frank–Wolfe method, Nesterov’s accelerated gradient method, ODE interpretations, dual methods, Nesterov’s smoothing, proximal gradient methods, Moreau–Yosida regularization | 09 |
| III | Operator splitting methods Augmented Lagrangian methods, alternating direction method of multipliers (ADMM), monotone operators, Douglas–Rachford splitting, primal and dual decomposition | 09 |
| IV | Stochastic and nonconvex optimization Dual averaging, Polyak–Juditsky averaging, stochastic variance reduced gradient (SVRG), Langevin dynamics, escaping saddle points, landscape of nonconvex problems, deep learning | 09 |
| V | Two Use Case of ML optimization Techniques | 04 |
Text Books
- Stephen Boyd and Lieven Vandenberghe’s book: Convex Optimization
- Nesterov’s old book: Introductory Lectures on Convex Optimization: A Basic Course
- Neal Parikh and Stephen Boyd’s monograph: Proximal Algorithms
- S´ebastien Bubeck’s monograph: Convex Optimization: Algorithms and Complexity
- Moritz Hardt’s Berkeley EE 227C course note
- Prateek Jain and Purushottam Kar’s survey on nonconvex optimization
- Linear Algebra and Learning from Data, Gilbert Strang
- Convex Optimisation by Stephen Boyd 10.Optimisation for Machine Learning by Suvrit Sra, MIT Press.
Ratings & Reviews
No reviews yet — share your experience with this item.