/Basic-NN-NumPy

Basic Neural Network example based on Simple Linear Regression (Supervised Machine Learning)

Primary LanguagePython

Basic-NN-NumPy

Basic Neural Network example based on Simple Linear Regression (Supervised Machine Learning)

Definitions:

Supervised Machine Learning Programming: We give the program our inputs (which is usually historical data), and our targets (the answers we aim at), and we expect from the program to give us the underlying rules. We will then use this model on new data.

image



Building Blocks of a Machine Learning Algorithm


A. Data --> Historical Data


B. Model --> E.g. Linear Regression


C. Objective Function --> "Loss Functions" or "Reward Functions"


D. Optimization Algorithm --> By varying the Model's parameters, the algorithm tries to find the minimum of the OF: Gradient Descent (SGD)




Elements of the Model in Supervised ML:


    

  1. Inputs (given by us)
    2. 

  2. Weights (expected from machine)
    3. 

  3. Biases (expected from machine)
    4. 

  4. Outputs (expected from machine)
    5. 

  5. Targets (given by us)
    6.