This course is designed to teach the student the concepts of supervised and unsupervised machine learning by experimenting on the toy datasets that are installed in Python's machine learning library, sklearn.
The student will learn the basics of coding in the Python programming language and then will learn the basics of machine learning by studying a small dataset and the code that has been used to make predictions on it.
The student will also be introduced to two different types of estimators, being the linear regression and logistic regression models. These two models are very commom and are instrumental in forming accurate predictions on test data.
When the student has learned the basics of programming in Python and making predictions on a very small dataset, he will go on to study the eight toy datasets that are installed in sklearn, which is Python's machine learning library. These toy datasets are:-
1. Boston house prices dataset
2. Iris flowers identification dataset
3. Diabetes regression dataset
4. Digits dataset
5. Linnerud dataset
6. Wine recognition dataset
7. Breast cancer dataset
The student will study the code of the above dataset and will learn the basics of supvervised machine learning, which involves making predictions on labeled datasets to answer either regression or classification problems.
The students will also go over the code on an unsupervised learning technique, clustering.
The process of making predictions on a dataset involves:-
1. create the program in Google Colab, which is a free online Jupyter Notebook hosted by Google.
2. Import the libraries that will be used to execute the program.
3. Load sklearn's toy datrasets into the program by using sklearn's method designed for this purpose.
4. Plot graphs of where the data points in the datasets reside in memory to analyse the data.
5. Scale, standardise or normalise the data.
6. Split the data into training and testing datasets.
7. Select a model from the sklearn library to use and make predictions on it, being either a classificier or a regressor.
8. Analyse the predictions by checking its accuracy or errors, and plotting a graph of the predictions versus the actual values.
9. Study clustering, which is a type of unsupervised learning.
