A list of data science resources.
| Resource | Comments |
|---|---|
| R for Data Science | Introduction to modern R programming using the tidyverse |
| Text Mining with R | Guide to a modern approach to text mining in R |
| The caret package | Guide to performing machine learning in R using the caret package |
| An Introduction to Statistical Learning | A book on machine learning with examples in R |
| Shiny | Learning for to use the Shiny package to create interactive dashboards |
| Advanced R | Advanced guide to R, particularly good is the style guide |
| R packages | Guide to writing packages |
| The reticulate package | Website of the reticulate package that allows you to use Python functions within R. For example, see this blog where it is used to embed a Python model within a Shiny web app. (Similarily look at the feather package for passing dataframes between R and Python.) |
| R-bloggers | Blogs about the use of R in analytics |
| knitr in a knutshell | A short introduction to the knitr package for reproducible research |
| Resource | Comments |
|---|---|
| bl.ocks | Website showing popular D3 examples |
| Observable | A notebook approach to written D3 and javascript |
| Search the Bl.ocks | Search D3 examples produced my others (great for inspiration!) |
| D3 Tips and Tricks | A good book about D3 |
| D3 in Depth | A good introduction to writing D3 |
| D3 tutorial list | A list of D3 tutorials from the D3 website |
| A better way to structure D3 code | Interesting blog post on how to strucutre D3 code |
| Eloquent JavaScript | Knowing a bit of JavaScript is a prerequisite for mastering D3 and this book is a good introduction |
| A Tour Through the Visualization Zoo | A good introduction to a wide range of visualisations you could do in D3 (though here they have been done in a precusor to D3) |
| dc.js | A library that combines D3 and crossfilter that makes it easier to create interactive dashboards |
| Resource | Comments |
|---|---|
| scikit-learn | sklearn is the go-to Python package for machine learning and the documentation is a worth of information, not only on usage but also about the techniques themselves |
| Modern Pandas | A guide to using pandas dataframes |
| imbalanced-learn | A package to deal with classifying imbalanced data with excellent documentation |
| Natural Language Processing with Python | This is a book on NLP in Python from the team behind the NLTK package. For text mining you should also look into spaCy and gensim (for topic modelling) |
| Requests: HTTP for Humans | Library for making HTTP requests from Python, great way of making API calls |
| Flask | Python framework for creating web apps |
| Seven Strategies for Optimizing Numerical Code | Slides on different approaches to speeding up Python code |
| Comparing Python Clustering Algorithms | Does what it says on the tin! |
| Style Guide for Python Code | This is PEP 8, the official style guide for Python. One incentive to following its guidance is that your code will better integrate with IDEs |
| Resource | Comments |
|---|---|
| Feature Engineering and Selection | Guide to feature engineering and model selection |
| Kaggle Ensembling Guide | A guide to combining models to approve performance |
| Elements of Statistical Learning | The classic text on machine learning |
| Resource | Comments |
|---|---|
| Neural Networks and Deep Learning | Simple introduction to neural networks |
| Convolutional Neural Networks for Visual Recognition | Stanford course on convolutional neural networks |
| Understanding Convolutional Neural Networks for NLP | Article explaining CNNs in the context of NLP |
| On word embeddings | Introduction to word embeddings |
| fast.ai | Online AI course |
| Resource | Comments |
|---|---|
| Towards Data Science | Interesting articles about data science |
| Data Science Weekly | Weekly data science newsletter that aggregates articles on data science |
| Why Use Make | Thoughts from Mike Bostock on using make for reproducible research |
| Statistical Modeling: The Two Cultures | Leo Breiman's article on the difference between statistical models and algorithmic models |