I'm going to solve some convex optimization problems in this repository as I'm going through my learning path of this topic.
Optimization problems are ubiquitous in statistics and machine learning. A huge number of problems that we consider in these disciplines can indeed be posed as optimization tasks.
Studying the details is important for two major reasons:
Half of the Machine Learning field is the construction of the learning method (optimization), there're many techniqques used for optimization if the optimization problem is solvable:
- Closed form solution ==> Least Mean squares
- Numerically and guaranteed ==> Conex optimization & Linear Programming
- Numerically but not guaranteed ==> Non-Convex, and can be solved using many techniques:
- Numerical algorithms like Gradient descent.
- Local optimization.
- Heuristic swarm-based methods.
- Brute force exhaustive search.