Traditional methods still relevant to big data?
Hi! Just a question about the end of the video where it mentions that traditional methods are better suited for traditional data, and ML is better suited for big data. Is this just a rule of thumb that doesn't always apply? For instance, I think of PCA as more of a 'traditional method', but it is still applied with high dimensional data that you might find in big data. Thanks for any input!
Hi Eleanor!
Thanks for reaching out.
Plenty of the traditional methods are used in ML today, at least according to the vocabulary used by many analysts. For instance, logistic regression is a very traditional method, but it can be applied also in a machine learning context. Therefore, depending on the context, we can talk about traditional methods/data or ML/big data.
The same principle applies for PCA which is, historically at least, a traditional statistical method.
Hope this helps.
Best,
Martin
