You can find most of these approaches in the internet as well. I tried reducing the libraries and moving parts used in the codes in order to maintain the forward-compatibility of the repository. Some of these codes are assignments are for ce959-1: Deep Learning at Sharif University of Technology
CE_SSE.py: Plotting states of Cross Entropy and SSE loss functions
Iris Data.py: Iris Data classification using Softmax
Linear_regression.py: Linear Regression algorithm only using numpy
Perceptron_Algorithm_AND.py: Perceptron algorithm - AND operator
Perceptron_Algorithm_OR.py: Perceptron algorithm - OR operator
Perceptron_Algorithm_XOR.py: Perceptron algorithm - XOR operator
Titanic_logistic_regression.py: Titanic data with Logistic Regression