Analog Mappings for Communication in Real Time.
In this project there is implementation of an iterative algorithm and deep learning model. The goal is to improve the iterative algorithm using deep learning.
In iter.py file you can find the implemetation of iterative algorithm.
to run this part please select the following parameters:
s - number of sampling points from the fX(x) distribution
n_s - numebr of sampling points from the fN(n) distribution
my_ftol - the tolarence for the GD algorithm
my_maxIter - maximum number of iteration for the GD
select - pre defined destributions and other parameters
In proj.py file you can find the implemetation of deep learning model.
pre_train - select 1 if you need to prefit the model, else select 0
pre_comb_model - select 1 if you need to combine models, else select 0
s - number of samples from the fX(x) distribution
n_s - number of samples from the fN(n) distribution
please select the number of nuerons in each layer NEURONS_LAYER_1, NEURONS_LAYER_2, NEURONS_LAYER_3, NEURONS_LAYER_4, NEURONS_LAYER_5, NEURONS_LAYER_6
select - pre defined destributions and other parameters
HistEst.py - if you are interested to use some other randomized distribution, please import the following: HistEst and use this instead of fX(x)
plot.py - please use plot to display graph from gathered data