MDFT Academy Quizzes

What should you do to improve the predictions?

You are building a regression model to predict the wind speed at any point on earth during any time of the day. You decide to train your model on latitude, longitude, time of day, and elevation features.

After training, the MAE = 35 km/h. You are not happy with this result and decide to perform data processing on the input features in an attempt to improve the training process.

Which processing steps might improve the quality of the predictions?

Choose all correct answers from the options below.

I'm sorry, your answers are not correct.

Bin and one-hot encode the latitude and longitude is correct
One-hot encode the time of day is incorrect. The time of day is a numeric feature and should not be one-hot encoded.
Sparse-vector encode the elevation is incorrect. The elevation is a numeric feature, sparse-vector encoding is for text data.
Normalize the time of day and elevation features is correct

2. Regression

What should you do to improve the predictions?