ML-practice

As I learn about new algorithms in machine learing I try to apply them on datasets for fun.

2 class classification

Tried to apply regularised logistic regression to predict outcomes of dataset from kaggle.
https://www.kaggle.com/msjaiclub/2classclassification

Fig 1:dataset 1 plot

Fig 2:dataset 2 plot

Looking at the plot it can be seen that a linear hypothesis cannot satisfy the dataset .
So a polynomial hypothesis is used with regularisation.
Applying the algorithm the following results are obtained.
Accuracy on dataset 1 : 97.03
Accuracy on dataset 2 : 88.23

Fig 3:Hypothesis curve

I welcome any suggestions or improvements to algorithm or implementation.