Demo_GradientDescent

Implemented a basic code for Gradient Descent as part of Siraj Rawal’s course Intro - The Math of Intelligence

Code referenced from : https://github.com/llSourcell/Intro_to_the_Math_of_intelligence

Gradient descent visualization

https://raw.githubusercontent.com/mattnedrich/GradientDescentExample/master/gradient_descent_example.gif

Dependencies

• numpy

  Use pip to install any dependencies.

Dataset used

• data.csv file of the form :

  x,y

  Example:

   distance cycled(miles), calories burned(KCal)
   32.502345269453031,        31.70700584656992
   55.142188413943821,        78.211518270799232
   ........          ,     .......

Usage

• Used hardcoded data.csv for data.

• To run:

  $ python3 demo.py
  Starting gradient descent -> initial_b = 0, initial_m = 0, initial_error = 5565.10783448
  Running...
  After 1000 iterations -> final_b = 0.116832206834, final_m = 1.4771976436, final_error = 112.5981953