Choose a dataset from Kaggle.com or any other repository. The dataset size must be not less than 20 Mbytes (uncompressed).

Note: The dataset must have missing values, noise and outliers

in python

1- Create a heat map of the correlation matrix that shows correlation coefficients among all the variables in the dataset. What are your observations?

2- Deduct some statistical results from the datasets (at least two results and discuss it in detail)

3- Perform the normality test for the data and graphically represent the results. Transform the data if not normally distributed.

4- Develop any two classification/clustering/Regression models based on your dataset type. Briefly describe the interpretation of each model.

5- Select one of the developed models and perform hyper-parameter tuning using best combination of model parameters.

6-Compare the optimized model with the initial model and indicate whether the results are statistically significant?