mvrck96/Coursework--Optimisation

Coursework on global optimization methods (BGD, Adadelta)

BGD and Adadelta methods for function approximation problem

Short info

Implemented methods:

• Batch Gradient Descent
• Adadelta

Function to approximate: F(w, x): w_1 + w_2x + w_3x^2

Loss function: MSE

Further details could be found in paper.pdf

Requirements

• numpy
• matplotlib

References

1. Matthew D. Zeiler, ADADELTA: AN ADAPTIVE LEARNING RATE METHOD.
2. Sebastian Ruder, An overview of gradient descent optimization algorithms.