Icon

Solution 3 Training the Logistic Regression Algorithm

This workflow shows a solution to a hands-on exercise in the L4-ML Introduction to Machine Learning Algorithms self-paced course

Task 2: Compare the performances of the logistic regression and decision treealgorithms1. Predict the "class" column using the logistic regression algorithm. Use thedefault configuration.2. If the algorithm doesn't converge, try Gauss/Laplace regularization3. Apply the model to the test set and score its performance4. Compare the overall accuracy of the logistic regression model to the overallaccuracy of the decision tree model. Task 1: Inspect the class separation in the feature space 1. Visualize feature 0 vs feature 3 in a scatter plot color by classCheck class separationsplit intotraining and testReaddec-tree-data.tableonly features0,3 Color Manager Scatter Plot Partitioning DecisionTree Learner Decision TreePredictor Scorer Scorer Table Reader Column Filter LogisticRegression Learner Logistic RegressionPredictor Task 2: Compare the performances of the logistic regression and decision treealgorithms1. Predict the "class" column using the logistic regression algorithm. Use thedefault configuration.2. If the algorithm doesn't converge, try Gauss/Laplace regularization3. Apply the model to the test set and score its performance4. Compare the overall accuracy of the logistic regression model to the overallaccuracy of the decision tree model. Task 1: Inspect the class separation in the feature space 1. Visualize feature 0 vs feature 3 in a scatter plot color by classCheck class separationsplit intotraining and testReaddec-tree-data.tableonly features0,3Color Manager Scatter Plot Partitioning DecisionTree Learner Decision TreePredictor Scorer Scorer Table Reader Column Filter LogisticRegression Learner Logistic RegressionPredictor

Nodes

Extensions

Links