ml_steps.ipynb model_baseline requires self?

[i-namekawa]

In my installation (I did not use Docker as instructed, but I used Ubuntu 18.04 Anaconda Python 3.7.2), the jupyter notebook model_baseline did not make prediction due to an error expecting two args for the line below:

y_predict_baseline = model_v1.predict(df[['r', 'g', 'b']])

If I put self in the definition of predict function, the error was fixed.

 `def predict(self, X: pd.DataFrame) -> np.array:`