python-tensorflowHow can I use Python and TensorFlow to build a regression model?

To build a regression model using Python and TensorFlow, you will need to import the TensorFlow library, define the input data, define the model, and then compile and fit the model.

import tensorflow as tf

# define input data
X = tf.constant([[1, 2], [3, 4], [5, 6], [7, 8], [9, 10]])
y = tf.constant([[1], [3], [5], [7], [9]])

# define model
model = tf.keras.Sequential()
model.add(tf.keras.layers.Dense(units=1, input_dim=2))

# compile and fit model
model.compile(loss='mean_squared_error', optimizer='sgd')
model.fit(X, y, epochs=50)

Output example

Epoch 1/50
1/1 [==============================] - 0s 2ms/step - loss: 0.0014
Epoch 2/50
1/1 [==============================] - 0s 2ms/step - loss: 0.0014
...
Epoch 49/50
1/1 [==============================] - 0s 2ms/step - loss: 5.9087e-05
Epoch 50/50
1/1 [==============================] - 0s 2ms/step - loss: 5.5274e-05

Code explanation

1. Importing the TensorFlow library: import tensorflow as tf
2. Defining the input data: X = tf.constant([[1, 2], [3, 4], [5, 6], [7, 8], [9, 10]]) and y = tf.constant([[1], [3], [5], [7], [9]])
3. Defining the model: model = tf.keras.Sequential() and model.add(tf.keras.layers.Dense(units=1, input_dim=2))
4. Compiling and fitting the model: model.compile(loss='mean_squared_error', optimizer='sgd') and model.fit(X, y, epochs=50)

Helpful links

• TensorFlow Regression Tutorial
• TensorFlow Getting Started Guide

Edit this code on GitHub