Stay organized with collections
Save and categorize content based on your preferences.
Neural Net Spiral
This data set is a noisy spiral. Obviously, a linear model will fail here,
but even manually defined feature crosses may be hard to construct.
Task 1: Train the best model you can, using just X1 and
X2. Feel free to add or remove layers and neurons, change
learning settings like learning rate, regularization rate, and
batch size. What is the best test loss you can get? How smooth is
the model output surface?
Task 2: Even with Neural Nets, some amount of feature engineering is
often needed to achieve best performance. Try adding in additional
cross product features or other transformations like
sin(X1) and sin(X2). Do you get a better
model? Is the model output surface any smoother?
(Answers appear just below the exercise.)
Click the plus icon for possible answers.
The following video walks through how to choose hyperparameters in Playground
to train a model for the spiral data that minimizes test loss.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2024-08-21 UTC."],[[["The dataset presents a challenge for linear models due to its spiral nature, requiring neural networks for effective modeling."],["Task 1 focuses on achieving the lowest test loss using only the original features (X~1~, X~2~) by adjusting network architecture and learning settings."],["Task 2 explores improving model performance and output smoothness through feature engineering, such as adding cross product features or trigonometric transformations."],["The exercise provides an opportunity to experiment with hyperparameter tuning in neural networks for optimal results on the spiral dataset."]]],[]]