In the previous dataset, the were strongly correlated with the labels. However, in some time series, we have features that might be less correlated or not correlated with the labels at all. The main idea behind machine learning is that the algorithm tries to figure out by itself which features are valuable and which are not. Especially in deep learning, we want to keep the feature engineering limited. In the following recipe, we will be predicting the demand for bike sharing rentals. The data includes some interesting features, such as weather type, holiday, temperature, and season.
- First, we all libraries:
from sklearn import preprocessing import pandas as pd import numpy as np from math import pi, sin, cos from datetime import datetime from keras.models import Sequential from keras.layers import Dense, Activation, Dropout from keras.optimizers import Adam from keras.callbacks import EarlyStopping
- The training and test data is stored in two...