Machine learning isn’t magic — it’s a process. This section walks through the core workflow: exploring and shaping data, building and training models, and testing how well they perform. Expect practical tools, clear examples, and real evaluation — not black boxes or buzzwords.
It’s not the model — it’s what you do with it.
Don’t get stuck in a loop of adjusting sliders and chasing score decimals. Build projects. Focus on data (more data, cleaner data, feature engineering) for insight. Make something that solves a problem or tells a story.
Machine Learning Topics
- Explorator Data Analysis (EDA): Understand data properties, data types, missing data, outliers, and more key factors to effective ML.
- Classification: Models that work on categorized predictions (spam/not-spam).
- Regression: Models that prediction continuous values (stock prices).
- Evaluation Metrics: Determine how to evaluate a model and assess performance.
Need a place to start? Check out the FAQs!
Models & Metrics
Classification
Regression
I used to run dozens of models, hyperparameter tuning, and comparitive metrics. While it taught me a lot, the biggest lesson was:
Modeling is a tool, meaning is the goal.
Understand data — no good models are built on bad data
Fit Models — no good inference comes from inappropriate models
Test Performance — all models are poor until proven otherwise
“I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times”— Bruce Lee
Machine learning isn’t magic, and it’s not about finding the model — it’s about applying the right tools to meaningful questions. The most advanced models require on proper input and programming.