How I structured my Machine Learning Study

I have completed four weeks of my Data Science journey. In the last week, I went head first into Machine Learning. Machine Learning is a broad subject which falls under Artificial Intelligence. You know AI, right? When you request an Uber, when you open google maps to avoid high traffic or when you group your friend’s pictures in your photos app, are all instances of when you use AI applications. You may not know the definition but now you should have an idea of what I’m talking about.

Machine Learning falls under Artificial Intelligence and it is concerned with making predictions to make life easier for us. Have you ever heard the joke of how people say, “They are watching/listening to us?” Whoever ‘they’ is, I’m not sure. But that ideology is true. You shop or add to your wish list on Amazon or Instagram for Hair bundles and a day or two after you start to see hair products, wigs etc. appearing on your feed. Machine Learning (ML) algorithms are in place here to monitor user activity and predict based on their interests.

A tweet I posted about my Machine Learning study

There are many ML algorithms (a process followed to solve a problem) which fall under supervised and unsupervised learning categories. I had to determine how I was going to understand it all in one week. I don’t mean to sound pessimistic but it is close to impossible to understand all these algorithms at a deep level and get valuable practice experience at the same time. Learning takes time and the more you work on something, the better you get at it.

This is what I did. I had to study Classification, Linear Regression, Logistic Regression, Gradient Boosted Algorithms and Decision Tree Algorithms. I sought advice from people who already had a fair idea of ML to know how I should progress and which ones I needed to know first. Machine Learning is a discipline found at the intersection of Statistics, Mathematics and Computer Programming. I decided it was best to know the foundation of these three. I’ve studied the above subjects in school but I needed to know the statistics behind these algorithms I was meant to study. In my opinion Machine Learning is merely statistics translated from paper to a computer. Math is more or less the basis for statistics and when we want to create a prediction model, a computer makes that easier to realize.

I have a 30-minute interval weekly planner I use to schedule my tasks. I started with introductory reading and video lessons and moved on to the detailed ones. I began with Introduction to Machine learning, then to linear regression and then logistic regression. Decision tree algorithm would follow and then I would cover boosted algorithms as this stands on the concept of decision trees (I am yet to cover those). I have given access to anyone who wishes to download the weekly planner.

Machine Learning has its own biases and comes with its associated errors. Cross-validation helps to test one or more models multiple times to ensure its accuracy. On a side note, my German lessons have been sinking from inadequate attention :( . Aber es wird gut. Thank You.

Bis bald!

Important resources for further learning:

Applications of machine learning from day to day life

How machine learning can help your business

Sentiment analysis