You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi all, this is a very nice chart but I believe that there should be slight modifications.
Data Science Roadmap
"Dimensionality and Numerosity Reduction" is a large topic which would include the study of the "Principal Component Analysis" (PCA) algorithm as the very first thing you'd do. Yet PCA is seen as being the very final thing you look at in this section. I think you should put PCA right before Dimensionality Reduction or you should combine them.
In the "Visualization section" you list some nice plotting libraries, but I think that Bokeh should be included here because it is a superior python plotting library (out of the box GPU/OpenGL support allowing for plotting millions of points, significantly more flexible interaction system) combined to the other options and is quickly becoming one of the important graphing libraries.
Under "Data Sources" you may want to put "Data Mining and Web Scraping" or something along those lines since I think a Data Scientist should be able to get their own data rather than go on Kaggle or github awesome pages.
Machine Learning Roadmap
Subsections under "Association Rule Learning" should be "Apriori algorithm", "ECLAT algorithm" and "Fp-trees"
Subsections under Dimensionality Reduction should include (after PCA): "Random Projection", "NMF", "T-SNE", "UMAP"
Subsections under Clustering should include (after Agglomerative): "OPTICS" and "HDBSCAN"
Subsections under "Classification" should include "Guassian Mixture Models"
Logistic Regression is actually a binary classification algorithm, despite its name, so move it from regression to Classification
Moving Huggingface Transformers out from here and into the Deep Learning section.
Deep Learning Roadmap
Add a new section under "Architectures" called "Attention Mechanisms/Transformers"
Add a new section under Architectures called "NEAT/Evolving Architectures"
Big Data Engineer
Add a new blue section under "Tools" for Dask, Numba, Onnx, and OpenVino
If it's really easy to generate these plots, I'm willing to make these changes and submit a PR. What are your thoughts on implementing some or all of these changes?
The text was updated successfully, but these errors were encountered:
Hi all, this is a very nice chart but I believe that there should be slight modifications.
Data Science Roadmap
"Dimensionality and Numerosity Reduction" is a large topic which would include the study of the "Principal Component Analysis" (PCA) algorithm as the very first thing you'd do. Yet PCA is seen as being the very final thing you look at in this section. I think you should put PCA right before Dimensionality Reduction or you should combine them.
In the "Visualization section" you list some nice plotting libraries, but I think that Bokeh should be included here because it is a superior python plotting library (out of the box GPU/OpenGL support allowing for plotting millions of points, significantly more flexible interaction system) combined to the other options and is quickly becoming one of the important graphing libraries.
Under "Data Sources" you may want to put "Data Mining and Web Scraping" or something along those lines since I think a Data Scientist should be able to get their own data rather than go on Kaggle or github awesome pages.
Machine Learning Roadmap
Subsections under "Association Rule Learning" should be "Apriori algorithm", "ECLAT algorithm" and "Fp-trees"
Subsections under Dimensionality Reduction should include (after PCA): "Random Projection", "NMF", "T-SNE", "UMAP"
Subsections under Clustering should include (after Agglomerative): "OPTICS" and "HDBSCAN"
Subsections under "Classification" should include "Guassian Mixture Models"
Logistic Regression is actually a binary classification algorithm, despite its name, so move it from regression to Classification
Moving Huggingface Transformers out from here and into the Deep Learning section.
Deep Learning Roadmap
Add a new section under "Architectures" called "Attention Mechanisms/Transformers"
Add a new section under Architectures called "NEAT/Evolving Architectures"
Big Data Engineer
Add a new blue section under "Tools" for Dask, Numba, Onnx, and OpenVino
If it's really easy to generate these plots, I'm willing to make these changes and submit a PR. What are your thoughts on implementing some or all of these changes?
The text was updated successfully, but these errors were encountered: