# S-Rocket: Selective Random Convolution Kernels for Time Series Classification

arXiv: https://arxiv.org/abs/2203.03445 · HTML: https://mlsanity.com/paper/2203.03445/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 34 of its 65 references that are in the map.
- First posted on arXiv 7 Mar 2022.

## Builds on (most cited first, 30 of 34)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling](https://mlsanity.com/paper/1803.01271/) (2018)
- [How transferable are features in deep neural networks?](https://mlsanity.com/paper/1411.1792/) (2014)
- [Pruning Filters for Efficient ConvNets](https://mlsanity.com/paper/1608.08710/) (2016)
- [SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size](https://mlsanity.com/paper/1602.07360/) (2016)
- [Network In Network](https://mlsanity.com/paper/1312.4400/) (2013)
- [What is the State of Neural Network Pruning?](https://mlsanity.com/paper/2003.03033/) (2020)
- [Deep learning for time series classification: a review](https://mlsanity.com/paper/1809.04356/) (2018)
- [Recurrent Neural Networks for Multivariate Time Series with Missing Values](https://mlsanity.com/paper/1606.01865/) (2016)
- [The UCR Time Series Archive](https://mlsanity.com/paper/1810.07758/) (2018)
- [InceptionTime: Finding AlexNet for Time Series Classification](https://mlsanity.com/paper/1909.04939/) (2019)
- [ROCKET: Exceptionally fast and accurate time series classification using random convolutional kernels](https://mlsanity.com/paper/1910.13051/) (2019)
- [Soft Filter Pruning for Accelerating Deep Convolutional Neural Networks](https://mlsanity.com/paper/1808.06866/) (2018)
- [A Survey of Model Compression and Acceleration for Deep Neural Networks](https://mlsanity.com/paper/1710.09282/) (2017)
- [Scalable and accurate deep learning for electronic health records](https://mlsanity.com/paper/1801.07860/) (2018)
- [MINIROCKET: A Very Fast (Almost) Deterministic Transform for Time Series Classification](https://mlsanity.com/paper/2012.08791/) (2020)
- [Towards moderate overparameterization: global convergence guarantees for training shallow neural networks](https://mlsanity.com/paper/1902.04674/) (2019)
- [Structured Pruning of Deep Convolutional Neural Networks](https://mlsanity.com/paper/1512.08571/) (2015)
- [Recent Advances in Recurrent Neural Networks](https://mlsanity.com/paper/1801.01078/) (2017)
- [Proximity Forest: An effective and scalable distance-based classifier for time series](https://mlsanity.com/paper/1808.10594/) (2018)
- [Resource-Efficient Neural Networks for Embedded Systems](https://mlsanity.com/paper/2001.03048/) (2020)
- [EDropout: Energy-Based Dropout and Pruning of Deep Neural Networks](https://mlsanity.com/paper/2006.04270/) (2020)
- [MINT: Deep Network Compression via Mutual Information-based Neuron Trimming](https://mlsanity.com/paper/2003.08472/) (2020)
- [Extreme Classification via Adversarial Softmax Approximation](https://mlsanity.com/paper/2002.06298/) (2020)
- [Multi-Objective Pruning for CNNs Using Genetic Algorithm](https://mlsanity.com/paper/1906.00399/) (2019)
- [LiteHAR: Lightweight Human Activity Recognition from WiFi Signals with Random Convolution Kernels](https://mlsanity.com/paper/2201.09310/) (2022)
- [Ising-Dropout: A Regularization Method for Training and Compression of Deep Neural Networks](https://mlsanity.com/paper/1902.08673/) (2019)
- [G-softmax: Improving Intra-class Compactness and Inter-class Separability of Features](https://mlsanity.com/paper/1904.04317/) (2019)
- [Deep Sequential Learning for Cervical Spine Fracture Detection on Computed Tomography Imaging](https://mlsanity.com/paper/2010.13336/) (2020)
- [Pruning of Convolutional Neural Networks Using Ising Energy Model](https://mlsanity.com/paper/2102.05437/) (2021)

## Built on by (most cited first, 7 of 7)

- [Detach-ROCKET: Sequential feature selection for time series classification with random convolutional kernels](https://mlsanity.com/paper/2309.14518/) (2023)
- [POCKET: Pruning Random Convolution Kernels for Time Series Classification from a Feature Selection Perspective](https://mlsanity.com/paper/2309.08499/) (2023)
- [Hypertension Detection From High-Dimensional Representation of Photoplethysmogram Signals](https://mlsanity.com/paper/2308.02425/) (2023)
- [Time series classification with random convolution kernels: pooling operators and input representations matter](https://mlsanity.com/paper/2409.01115/) (2024)
- [HIERVAR: A Hierarchical Feature Selection Method for Time Series Analysis](https://mlsanity.com/paper/2407.16048/) (2024)
- [Pruning Extensions and Efficiency Trade-Offs for Sustainable Time Series Classification](https://mlsanity.com/paper/2604.07953/) (2026)
- [Random Feature Pruning for Multi-Class Classification](https://mlsanity.com/paper/2610.15101/) (2026)

## Abstract (arXiv)

Random convolution kernel transform (Rocket) is a fast, efficient, and novel approach for time series feature extraction using a large number of independent randomly initialized 1-D convolution kernels of different configurations. The output of the convolution operation on each time series is represented by a partial positive value (PPV). A concatenation of PPVs from all kernels is the input feature vector to a Ridge regression classifier. Unlike typical deep learning models, the kernels are not trained and there is no weighted/trainable connection between kernels or concatenated features and the classifier. Since these kernels are generated randomly, a portion of these kernels may not positively contribute in performance of the model. Hence, selection of the most important kernels and pruning the redundant and less important ones is necessary to reduce computational complexity and accelerate inference of Rocket for applications on the edge devices. Selection of these kernels is a combinatorial optimization problem. In this paper, we propose a scheme for selecting these kernels while maintaining the classification performance. First, the original model is pre-trained at full capacity. Then, a population of binary candidate state vectors is initialized where each element of a vector represents the active/inactive status of a kernel. A population-based optimization algorithm evolves the population in order to find a best state vector which minimizes the number of active kernels while maximizing the accuracy of the classifier. This activation function is a linear combination of the total number of active kernels and the classification accuracy of the pre-trained classifier with the active kernels. Finally, the selected kernels in the best state vector are utilized to train the Ridge regression classifier with the selected kernels.