After a fair bit of conversing with GPT 3.5 about unsupervised & supervised learning I have to try to build use cases of unsupervised learning over real world data.
Practical machine learning starts with data mining. And for data mining we need data collected neatly(close to neat) This has been done in various projects throughout the years. Needless to say its all real world data.
Approaching ML from a SDE point of view, taking baby steps, reading think-stats, understanding histograms, distributions, quantiles, normalisation, encoding, scikit-learn, scipy avoiding using neural nets till statistical models arent understood and worked upon. nlp/embedding related ML notebooks not included in this folder. That is a different subject as of now.
lots of link in notebooks. Important to read all in there as they are the filtered, top ones from a large pile of reading material.
- Notebooks with Kaggle datasets
- Stochaistic process