Introduction to Machine Learning
Machine Learning Series
When I first started independently working on machine learning problems, I found some difficulty with knowing which algorithms to choose or which metrics would best evaluate model performance. My university education had taught me these things, but the learning curve for my data mining class was famously steep. Somewhere outside the comfort of the classroom, I had lost my calm and become marooned in information overload.
With the cacophonous hype, daunting terminology and unrelenting expectations on even the most junior data scientists, my situation was taxing but far from unique.
I have always trusted myself to develop a solution-oriented mindset to approach problems. Doing this requires curiosity and the will to overcome mental resistance to compartmentalise seemingly overwhelming amounts of information into small, clear patterns. Most (if not all) complex things in life exist as layer upon layer of simpler concepts, and examining the individual building blocks allows us to gain perspective.
Knowing this, I went back to the basics. If I wanted to be a skilled data science practitioner, I would have to be comfortable with rebuilding my understanding of the nuances of decision-making in machine learning.
If you find or have ever found yourself in a similar boat to mine, I hope this series helps you. We will be working with multiple datasets to build familiarity with different machine learning problems.
Resources for ML Series:
