Machine Learning 101: Linear Regression Simplified

From Hackathons to Horizons -- The Beginning: 2018 – A Hackathon, A Vision

It all began in 2018 (for me), during what seemed like just another hackathon. But for me, it was the spark that ignited everything.

I had the honor of leading a small but fierce team called Prognosis Pundits(Jayesh, Akash, Rohit and others). Our mission? To solve a seemingly simple yet deeply impactful problem: automated resume ranking. At the time, it felt ambitious. We poured days (and nights) into feature engineering, model selection, and UX design—testing, failing, refining. And then, it happened: we won.

That victory wasn’t just a trophy or a title. It was a validation of our ideas, our teamwork, and our belief in what AI could do to solve real-world problems.

It all starts with the most basis yet defining concept. One of the most essential  concepts to be clarified before embarking on the journey of machine learning was linear regression. It is a simple concept both mathematically and to visualize practically. However still when I came across below video I thought a 5th grade kid could also understand linear regression. :)
So here is the video:

Introduction to Linear Regression

Need to always start with basic questions:

What is regression? 
Regression: Modelling target value using independent predictors. Helps in forecasting and finding out cause and effect relationship between variables. Regression techniques differ based on the types of relationship between dependent and independent variable and the number of independent variables.

Linear regression: The number of independent variables is one and there is a linear relationship between the independent(x) and dependent(y) variable.

and then we continued to build on using the data set, mentors and guides provided to us. 

Learning, Unlearning, Relearning of Machine Learning

While we pivoted back to our routine(assigned) business projects where the scale and scope both did not have a use case for AI, for the industry, what followed was a whirlwind of growth. From shallow classifiers to deep neural networks, from handcrafted features to transfer learning and transformers—the field evolved.

During this immersive hackathon, we learned that being good at AI wasn’t just about models and math. It was about curiosity, empathy, and humility. It was about listening to users, interpreting data with context, and always questioning the “why” behind the “what.”

We made mistakes—plenty of them. But every bug, every failed experiment, every overfit model taught us more than any textbook ever could.

The People, the Process, the Purpose

Gratitude, above all, goes to the people I’ve met along the way—mentors who challenged us, teammates who inspired me, and the broader AI community that has open arms and sharp minds.

Each collaboration added a new layer to my understanding—not just of technology, but of humanity. Because at the end of the day, the best AI systems are built not just with data and code, but with purpose and people at the core.