35: Logistic Regression | TensorFlow | Core API | Tutorial смотреть онлайн

The video discusses in TensorFlow core API: Logistic regression for binary classification
Dataset: Breast Cancer

00:00:00 - Overview
00:00:37 - Import libraries
00:02:27 - Set figure sizes
00:04:25 - Load dataset: Wisconsin Breast Cancer
00:10:36 - Check data: dataset.info(), .unique()
00:11:33 - Split dataset into train and test: .sample(), .drop()
00:13:19 - Separate features and target: x_train, y_train, x_test, y_test
00:16:25 - Preprocessing: .map()
00:19:54 - Preprocessing: tf.convert_to_tensor()
00:22:23 - Preprocessing: sns.pairplot()
00:24:57 - Descriptive statistics: .describe()
00:26:31 - Normalization: create class Normalize(tf.Module)
00:39:13 - * * * ERROR mentioned at 01:31:48 * * *:
00:29:07 - Normalization: normalize x_train, x_test
00:30:39 - Create and visualize sigmoid function
00:34:30 - Create Log loss function
00:35:50 - Gradient descent update rule: create class LogisticRegression(tf.Module)
00:39:13 - * * * ERROR mentioned at 01:13:39 * * *: self.built=True
00:41:13 - Check if class LogisticRegression returns values in range 0 to 1
00:44:32 - Create accuracy function
00:48:30 - Convert data: tf.data.Dataset.from_tensor_slices()
00:51:00 - Convert data: .shuffle().batch()
00:53:00 - Training: Set training parameters
01:05:59 - * * * ERROR * * *: test_losses.append(test_loss), test_accs.append(test_acc)
01:08:47 - Performance evaluation: plots for loss and accuracy
01:11:07 - * * * ERROR * * * (x * self.std) + self.mean
01:13:39 - * * * ERROR * * * self.built=True
01:17:00 - Performance evaluation: Confusion matrix
01:24:17 - Save model: create class ExportModule(tf.Module)
01:29:41 - Load saved model
01:31:24 - Ending notes

# ----------------
# TensorFlow Guide
# ----------------
https://www.tensorflow.org/guide/core/logistic_regression_core