33: Linear Regression | TensorFlow | Tutorial

The video discusses in TensorFlow: Regression to predict vehicle fuel efficiency using Auto MPG dataset. 00:00:00 - Overview 00:01:56 - Import libraries 00:03:20 - Download data: Auto MPG 00:06:14 - Data: .head(), .dtypes, .isna().sum() 00:07:51 - Data: .map() 00:09:31 - Data: create dummies: pd.get_dummies() 00:11:31 - Data: split train and test: .sample(frac=0.8) 00:13:00 - Data: inspect or explore: sns.pairplot() 00:15:27 - Data: statistics: .describe() 00:16:44 - Data: separate features and target: .pop() 00:18:00 - Normalization: tf.keras.layers.Normalization() 00:20:25 - Normalization: .adapt() 00:24:50 - Linear Regression: One variable 00:26:16 - Linear Regression: One variable: Normalization: layers.Normalization() 00:27:04 - Linear Regression: One variable: Build model: tf.keras.Sequential() 00:28:14 - Linear Regression: One variable: Predict: .predict() [without training] 00:29:19 - Linear Regression: One variable: Compile: .compile() 00:30:33 - Linear Regression: One variable: Fit: .fit() 00:32:33 - Linear Regression: One variable: validation loss 00:33:20 - Create function to plot loss 00:34:55 - Linear Regression: One variable: plot loss 00:36:02 - Linear Regression: One variable: Test results, plots 00:40:30 - Linear Regression: Multiple variables: tf.keras.Sequential() 00:41:38 - Linear Regression: Multiple variables: Predict: .predict() [without training] 00:42:30 - Linear Regression: Multiple variables: .kernel 00:43:03 - Linear Regression: Multiple variables: .compile() 00:43:49 - Linear Regression: Multiple variables: .fit() 00:44:44 - Linear Regression: Multiple variables: plot loss 00:45:30 - Linear Regression: Multiple variables: test: .evaluate() 00:46:31 - Linear Regression: Deep Neural Network (DNN): function to create model 00:49:06 - Linear Regression: Deep Neural Network (DNN): One variable: build, compile 00:51:00 - Linear Regression: Deep Neural Network (DNN): One variable: .fit() 00:51:57 - Linear Regression: Deep Neural Network (DNN): One variable: plot loss 00:53:37 - Linear Regression: Deep Neural Network (DNN): One variable: Test: .evaluate() 00:54:48 - Linear Regression: Deep Neural Network (DNN): Multiple variables: build, compile 00:56:10 - Linear Regression: Deep Neural Network (DNN): Multiple variables: .fit() 00:56:33 - Linear Regression: Deep Neural Network (DNN): Multiple variables: plot loss 00:57:30 - Performance: MAE 00:58:40 - Make predictions: .predict() 00:59:33 - Plot true values vs. predictions 01:01:38 - Error distribution: Plot 01:04:07 - Save model: .save() 01:04:30 - Reload saved model: tf.keras.models.load_model() 01:07:20 - Ending notes # ---------------- # TensorFlow Guide # ---------------- https://www.tensorflow.org/tutorials/keras/regression