The learner will open a go into the Kaggle website and open their own account. He will go into the competitions page of Kaggle and will enter five Kaggle competitions. He will review the code of the following Kaggle competitions:-
1. Titanic competition
2. Ames House Price competition
3. Disaster Tweet competition
4 Recognise the Digit competition
5. Predict Future Sales competition.
When the learner reviews the code, he will learn a methodical procedure for writing code onto Kaggle's Jupyter Notebook for the competition. He learner will learn how to accomplish the following tasks when he enters the competitions:-
1. He will learn how to import libraries into the Jupyter Notebook.
2. The learner will learn the purpose of the libraries he is importing, to include pandas, numpy, os, sklearn, math, matplotlib, seaborn, nltk, and string, just to name a few.
3. The learner will learn how to read a csv file into the program by using the pandas library to convert it into a dataframe.
4. The learner will learn how to analyse the csv and determine what type of features it consists of.
5. The learner will learn how to analyse the target and use matplotlib or seaborn to graph the data points.
6. The learner will learn how to clean up code and replace missing values.
7. The learner will learn how to combine two dataframes together to form one dataframe.
8. The learner will learn the difference between numerical and categorical values, and will learn how to encode categorical values.
9. The learner will learn how to make a heatmap to carry out feature selection of the model, thereby reducing potential noise in the system.
10. The learner will learn how to define features that will be used to make a prediction.
11. The learner will learn how to define the independent and dependent variables.
12. The learner will learn how to normalise and standardise the the independent variable because because predictions can be more accurate by using this preprocessing technique.
13. The learner will learn how to split the dataset into training and validatiog sets to be used in training, fitting and making predictions.
14. The learner will be able to use sklearn, a machine learning library, so select the most appropriate model to train, fit and predict on the data.
15. The learner will be able to make predictions on the test dataset, which will form the basis of the submission of predictions to Kaggle.
16. The learner will be able to prepare the predictions for submission and submit them to Kaggle for scoring.
17. Once the predictions have been submitted to Kaggle, the learner will be able to view his score and to look at his place on the leaderboard.
