to develop your data science skills
Jax is a Python library developed by Google in 2018 and is set to overtake Google’s other Python library, Tensorflow, for research purposes. There is significantly less code available in Jax than there is in Tensorflow, which is why I have decided to develop a course in Jax.
What you’ll learn
- Students will learn about Python’s Jax library..
- Students will learn how to code supervised classification machine learning algorithms in Jax..
- Students will learn how to code supervised regression machine learning algorithms in Jax..
- Students will learn how to code neural networks in Jax..
Course Content
- Introduction –> 2 lectures • 29min.
- Linear regression –> 4 lectures • 48min.
- Logistic regression –> 4 lectures • 43min.
- Naive Bayes –> 2 lectures • 16min.
- Decision tree –> 2 lectures • 17min.
- Random Forest –> 2 lectures • 16min.
- K Nearest Neighbour –> 2 lectures • 22min.
- Support Vector Machine –> 2 lectures • 15min.
- Neural Networks –> 6 lectures • 1hr 32min.
Requirements
Jax is a Python library developed by Google in 2018 and is set to overtake Google’s other Python library, Tensorflow, for research purposes. There is significantly less code available in Jax than there is in Tensorflow, which is why I have decided to develop a course in Jax.
Jax has been written very similar to the numpy API, but there are a few differences that will be covered in the course.
The beginning of the course will cover an introduction to Jax, discussing some of the code that will be in the 16 Jupyter Notebooks that will be presented. An introduction to machine learning algorithms will be vovered in eight sections. The machine learning algorithms that will be introduced, with the code covered in depth are:-
1. Linear regression
2. Logistic regression
3. Naive bayes
4. Decision tree
5. Random forest
6. K nearest neighbour
7. Support vector machine
8. Neural networks
In order for the machine learning algorithms to be efficiently presented, they must be included in a machine learning project, to include:-
1. Import Jax and other Python libraries into the program
2. Load the appropriate dataset into the program from Google Colab, GitHub, or sklearn
3. Preprocess the data if necessary
4. Remove outliers if appropriate
5. Remove highly correlated features if appropriate
6. Standardise the data if needed
7. Define dependent and independent variables
8. Split the dataset into training, validating, and testing sets, whichever is appropriate
9. Define the Jax model
10. Compare the Jax model with its sklearn equivalent
11. Obtain predictions and test their accuracy or error, whever is appropriate.