Build the simplest model

import numpy as np
import pandas as pd

Keras is the simplest ML library, so I hope you enjoy it

from keras.models import Sequential
from keras.layers import Dense

model = Sequential()

model.add(Dense(units=2, activation='relu', input_dim=2))
model.add(Dense(units=1))

model.compile(
    loss='mean_squared_error',
    optimizer='sgd',
    metrics=['accuracy']
)

1. first, we created a sequential model

2. then, we added two dense layer(actually neural network) for it. the first layer has one neural node, the second has one neural node, too.

3. sgd = Stochastic gradient descent

After that, let us feed data into our model

model.fit(x, y, epochs=100, batch_size=1)

Then do a prediction

model.predict(np.array([[170, 96]]))