In this academic project, we implemented and compared betwen regression methods for linear, multi-linear and exponential problems.
For the linear case, the implemented algorithms are : Newton method, Gradient Descent, Optimal Gradient Descent and Conjugant gradiant descent.
For the multilinear case, we implemented same methods applied to the linear case and added reguilarization to correct overfitting.
Finally, for th exponential case, the implemented methods are : Gradient Descent, Inetia method and Nestrov's accelerated Gradient Descent.