Lecturer
1. Noah Gift:
- a machine learning lecturer at Northwestern's Graduate Data Science program and UC Davis Graduate School of Management's MSBA program.
- the author of the book Pragmatic AI: An Introduction to Cloud-Based Machine Learning.
2. Kennedy Behrman:
- a professional consultant specializing in architecting and implementing cloud solutions for early-stage startups.
- in data engineering, data science, cloud solutions, engineering management, and have acted as a technical editor on numerous python and data science related publications.
Main Content
1. The history of python and data science.
2. Colab notebooks
3. Basic concepts in python
4. Strings and string formatting methods
5. Python data structures
6. Common data conversion recipes
7. execution control patterns
8. functions
9. common data science libraries: Tensorflow, Plotly, and seaborn.
10. functional programming
11. concept of lazy evaluation
12. pattern matching and text processing techniques
13. sorting in python and the creation of custom sort functions
14. core I/O concepts like concurrency and file handling
15. data scientist and Kaggle and GitHub ...
16. Case-study: apply real world data science problems in case studies
Lesson 1: Python past and future
- The Python language itself first appeared in 1990.
- ome of the goals of this course are to really talk about the future of Python in data science and also prepare new students for some of the things that are important and also ignore some things that are noT important. One of the things I'd like to discuss in the spirit of that is something that I call the three laws of automation. One, if it isn't automated, it's broken.
Two, be the automation, not the automated.
Three, if the automation of a task is being discussed it will eventually be automated.
Another thing that's really important is functions are everywhere. And this is really a core concept that is important to understand. It's the building block of data science, machine learning, and AI and it's about this input processing and output pipeline. If we look here, it would be we have this initial input here, there's the input, then we have the processing pipeline right here and then we have the output. And really, this goes on and on and on and there's millions and millions and millions of these types of building blocks that go everywhere inside of a distributed system that's doing deep learning and machine learning. So, it's really important to understand this functions are everywhere concept. Colab Notebooks are also a huge part of this lesson and every single lesson that we've talked about is in Colab Notebook. This is an emerging breakthrough technology. Jupyter, as I discussed, I've been actually writing about for over 12 years now and really Colab is the culmination of what this Jupyter and IPython technology really can be and it's the sharpest tool in the shed, so to speak. It's really the technology that is really going to allow you, as a newcomer to data science, to be the most productive. And what I like the most about Colab Notebooks is it's really about results over purity. So, on one hand, maybe you could have a more powerful ecosystem if you did everything yourself but from a results standpoint, Colab is just a powerful tool.