The dataset, make$\_$moons, has a lot of class imbalance and need to understand the benefits and assumptions underlying different resampling and rebalancing strategies.

$1.$ Create a set of experiments to explore how different rebalancing approaches work $($at least $3$ different methods$)$ for different levels of imbalance, and for different amounts of noise, variability, or signal structure.

$2.$ The dataset used in the code below has a specific assumption on its structure and noise, so it's important to find other ways to vary the data to understand how the methods will perform in different settings.

$3.$ Include visualizations $($i$.$e$.$ ROC curve, confusion matrix$)$ along with your code to demonstrate the findings.

$4.$ Discuss what experiments and research on the different resampling and reweighting strategies revealed.

The followings are the packages that can be included :

from imblearn.datasets import make$\_$imbalance

from collections import Counter

import matplotlib.pyplot as plt

import pandas as pd

from sklearn.datasets import make$\_$moons