# Insights into LSTM Fully Convolutional Networks for Time Series Classification

arXiv: https://arxiv.org/abs/1902.10756 · HTML: https://mlsanity.com/paper/1902.10756/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Feb 2019.

## Built on by (most cited first, 18 of 18)

- [LightTS: Lightweight Time Series Classification with Adaptive Ensemble Distillation -- Extended Version](https://mlsanity.com/paper/2302.12721/) (2023)
- [TE-ESN: Time Encoding Echo State Network for Prediction Based on Irregularly Sampled Time Series Data](https://mlsanity.com/paper/2105.00412/) (2021)
- [Data Augmentation for Time-Series Classification: An Extensive Empirical Study and Comprehensive Survey](https://mlsanity.com/paper/2310.10060/) (2023)
- [LLM-Mixer: Multiscale Mixing in LLMs for Time Series Forecasting](https://mlsanity.com/paper/2410.11674/) (2024)
- [Face Forgery Detection Based on Facial Region Displacement Trajectory Series](https://mlsanity.com/paper/2212.03678/) (2022)
- [LETS-GZSL: A Latent Embedding Model for Time Series Generalized Zero Shot Learning](https://mlsanity.com/paper/2207.12007/) (2022)
- [Multimodal Neural Network For Demand Forecasting](https://mlsanity.com/paper/2210.11502/) (2022)
- [Comparative Analysis of Deep Learning Models for Real-World ISP Network Traffic Forecasting](https://mlsanity.com/paper/2503.17410/) (2025)
- [Fault Detection and Diagnosis with Imbalanced and Noisy Data: A Hybrid Framework for Rotating Machinery](https://mlsanity.com/paper/2202.04212/) (2022)
- [Feature importance to explain multimodal prediction models. A clinical use case](https://mlsanity.com/paper/2404.18631/) (2024)
- [Condition Assessment of Stay Cables through Enhanced Time Series Classification Using a Deep Learning Approach](https://mlsanity.com/paper/2101.03701/) (2021)
- [Accuracy Improvement for Fully Convolutional Networks via Selective Augmentation with Applications to Electrocardiogram Data](https://mlsanity.com/paper/2104.12284/) (2021)
- [Improving Time Series Classification Algorithms Using Octave-Convolutional Layers](https://mlsanity.com/paper/2109.13696/) (2021)
- [TSAX is Trending](https://mlsanity.com/paper/2112.12912/) (2021)
- [On Principal Curve-Based Classifiers and Similarity-Based Selective Sampling in Time-Series](https://mlsanity.com/paper/2204.04620/) (2022)
- [TemporalAugmenter: An Ensemble Recurrent Based Deep Learning Approach for Signal Classification](https://mlsanity.com/paper/2401.06970/) (2024)
- [Statistical Analysis by Semiparametric Additive Regression and LSTM-FCN Based Hierarchical Classification for Computer Vision Quantification of Parkinsonian Bradykinesia](https://mlsanity.com/paper/2404.00670/) (2024)
- [Short-term Wind Speed Forecasting for Power Integration in Smart Grids based on Hybrid LSSVM-SVMD Method](https://mlsanity.com/paper/2408.17185/) (2024)

## Abstract (arXiv)

Long Short Term Memory Fully Convolutional Neural Networks (LSTM-FCN) and Attention LSTM-FCN (ALSTM-FCN) have shown to achieve state-of-the-art performance on the task of classifying time series signals on the old University of California-Riverside (UCR) time series repository. However, there has been no study on why LSTM-FCN and ALSTM-FCN perform well. In this paper, we perform a series of ablation tests (3627 experiments) on LSTM-FCN and ALSTM-FCN to provide a better understanding of the model and each of its sub-module. Results from the ablation tests on ALSTM-FCN and LSTM-FCN show that the LSTM and the FCN blocks perform better when applied in a conjoined manner. Two z-normalizing techniques, z-normalizing each sample independently and z-normalizing the whole dataset, are compared using a Wilcoxson signed-rank test to show a statistical difference in performance. In addition, we provide an understanding of the impact dimension shuffle has on LSTM-FCN by comparing its performance with LSTM-FCN when no dimension shuffle is applied. Finally, we demonstrate the performance of the LSTM-FCN when the LSTM block is replaced by a GRU, basic RNN, and Dense Block.