ML Foundations Course by Great Learning- Notes

In this blog, I am going to post my notes, assignments etc that I did during my course on ML Foundations by Great Learning.

ML Maths Basics

Topics Covered

• Line Concept
• Line, Planes and Hyper planes
• Vector Algebra-magnitude and dimension
• Vector Algebra-vector operations
• Dot Product
• Matrix Algebra
• Functions
• Maxima and Minima of Functions
• Chain Rule
• Maxima and Minima Applications in ML
• Gradient Descent using Partial Derivatives

Intro to AI and ML

AI-computer program that does something smart or makes smart decisions

When computer program learns about the world from data we call it ML.

We assume past is a good representation of future.

Model building from data

• take data as input
• find patterns in data
• summarise the pattern in a mathematically precise way

Machine Learning automates this model building.

If data is without noise then finding a pattern is easy but unfortunately data contains both data and noise. 

Noise is unstructured and random. It does not repeat itself.

ML does not assume data came from a specific model but statistics does.

ML tries all models to separate information from noise and find out which does better.

Most complicated model turns out to be that fits both information and data. This is called over fitting.

Very simple model manages to leave out information available to us which is called under fitting.

The goal is to find a balance and build a model that manages to capture as much information as it can and leave out noise. ML folks try a sequence of models.