Research map / Architectures and efficiency / RNNs and normalization
LassoNet: A Neural Network with Feature Sparsity
Ismael Lemhadri, Feng Ruan, Louis Abraham, Robert Tibshirani
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use LassoNet instead of Fisher score | feature selection (k=50) · MNIST-Fashion, NN learner | test accuracy: 0.671 -> 0.800 | Improved |
| use LassoNet instead of Fisher score | feature selection (k=50) · ISOLET, NN learner | test accuracy: 0.793 -> 0.885 | Improved |
| use LassoNet instead of PFA | feature selection (k=50) · Activity, NN learner | test accuracy: 0.779 -> 0.849 | Improved |
| use LassoNet instead of HSIC-Lasso | feature selection (k=50) · COIL-20, NN learner | test accuracy: 0.972 -> 0.991 | Improved |
| use LassoNet instead of PFA | feature selection (k=50) · MNIST, NN learner | test accuracy: 0.873 -> 0.873 | No clear effect |
| use LassoNet instead of HSIC-Lasso | feature selection (k=50) · Mice Protein, NN learner | test accuracy: 0.958 -> 0.958 | No clear effect |
| use LassoNet instead of PFA | feature selection (k=50) · MNIST, tree learner | test accuracy: 0.879 -> 0.892 | Improved |
| use LassoNet instead of HSIC-Lasso | feature selection (k=50) · ISOLET, tree learner | test accuracy: 0.888 -> 0.891 | No clear effect |
| use LassoNet instead of Fisher score | feature selection (k=50) · COIL-20, tree learner | test accuracy: 0.996 -> 0.993 | Hurt |
| use LassoNet instead of HSIC-Lasso | feature selection (k=50) · Activity, tree learner | test accuracy: 0.845 -> 0.860 | Improved |
| use 50 selected features instead of all features | LassoNet · MNIST, NN learner | test accuracy: 0.928 -> 0.873 | Hurt |
| use LassoNet instead of Soft-Impute for imputation | matrix completion · MICE Protein imputation | test MSE: about 30% reduction | Improved |
Closest papers by meaning
- The Contextual Lasso: Sparse Linear Models via Deep Neural Networks (NeurIPS 2023 · similarity 0.90)
- Sparsity May Cry: Let Us Fail (Current) Sparse Neural Networks Together! (ICLR 2023 · similarity 0.88)
- Sparsity in Deep Learning: Pruning and growth for efficient inference and training in neural networks (JMLR 2021 · similarity 0.88)
- Explainable Neural Networks with Guarantee: A Sparse Estimation Approach (AAAI 2025 · similarity 0.87)
- Learning Sparse Classifiers: Continuous and Mixed Integer Optimization Perspectives (JMLR 2021 · similarity 0.86)
- Exploring Structural Sparsity of Deep Networks Via Inverse Scale Spaces (TPAMI 2022 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
