typo error in notebook of excerise 6 of chapter 2 fixed

[sujal-GITHUB]

In the recent implementation of the StandardScalerClone class, we corrected a typo in Exercise 6 of Chapter 2 and enhanced its functionality to align with modern machine learning practices. The inverse_transform() method was added to ensure that executing scaler.inverse_transform(scaler.fit_transform(X)) returns results very close to the original input X. Additionally, the fit() method now populates the feature_names_in_ attribute with column names if the input X is a DataFrame, improving interpretability. The get_feature_names_out() method was also refined to accept an optional input_features parameter, validating its length and checking for matches with feature_names_in_ when defined, or returning a default NumPy array of feature names otherwise. These updates enhance the class's usability in preprocessing pipelines while resolving previous issues.