Normalize, Clean, Resample, and Train (NCRT): An End-to-End Pipeline Study for Solar Energetic Particle Forecasting

Abstract

Abstract Solar energetic particle (SEP) events are rare, high-impact space weather phenomena that pose serious radiation hazards to astronauts and polar-route aircraft crews, damage satellite electronics, and disrupt satellite communications and navigation systems. Forecasting SEPs allows preventive measures to be taken before an event's onset. This task is difficult because SEP events, defined as periods when the flux of > 10 MeV protons exceeds 10 proton flux units (pfu), are extremely rare and make up less than 1% of the samples in our dataset. We present a stage-by-stage study of a novel preprocessing pipeline for SEP forecasting (normalization -> borderline data cleaning -> class rebalancing -> training) and evaluate it across both classical and state-of-the-art classifiers (SVM, GRU, PatchTST, InceptionTime). To address data scarcity, we compare classical interpolation-based over-sampling (SMOTE, ADASYN) against two generative approaches, TimeGAN and a state-of-the-art diffusion model, for synthesizing minority-class sequences from only 118 real SEP events. Using the True Skill Statistic (TSS), we find that our novel hybrid log/z-score normalization and Tomek Links-based borderline data cleaning yield the largest gains, while classical over-sampling methods consistently outperform both generative approaches in this data-scarce regime. Our best configuration named NCRT-GRU achieves a TSS of 0.671 ± 0.00 on a test set that mimics an operational setting and outperforms recently published SEP forecasters. How to Use the Notebooks Run the notebooks in numeric order — each stage's output is the input to the next: 0_Dependancies.ipynb — installs required packages. Run once. 1_DataPreparation.ipynb — builds the multivariate time series dataset from the raw per-channel CSVs in dataset/, and produces the chronological 70/10/20 train/validation/test split used by every downstream experiment. 2_ImpactsOfNormalization.ipynb / 2.1_InitialClassification.ipynb — evaluate the four normalization schemes (Hybrid, Z-score, Min-Max, Log) and select classifier hyperparameters. 3_BorderlineCleaning.ipynb — applies Tomek Links, ENN, and NearMiss-3 to the best-normalized training set and compares results. 4_RandomUnderSampling.ipynb — evaluates RUS alone, at several retained-NSEP levels, on the cleaned training set. 5_RUS&OS-SMOTE.ipynb, 6_RUS&OS-TimeGAN.ipynb / 7_RUS&OS-TimeGAN-Results.ipynb, 8_RUS&OS-Adasyn.ipynb, 9_RUS&OS-Diffusion.ipynb / 10_RUS&OS-Diffusion-Results.ipynb — combine RUS with each of the four over-sampling methods (classical: SMOTE, ADASYN; generative: TimeGAN, diffusion) at every shared configuration, and report classification results. 11–19 (*_Figures_*.ipynb) — regenerate every figure in the paper from the saved experiment results. Run 11_Figures_Style.ipynb first in the same kernel session (it defines the shared color palette and result-loading helpers the later figure notebooks depend on), then run any of 12–19 as needed. Each notebook documents its own expected inputs/outputs in its first markdown cell. Repository Structure ├── dataset/ # Raw per-channel OMNI/GOES time series (gzipped CSVs) │ ├── F.csv.gz, Np.csv.gz, P4.csv.gz, P5.csv.gz, P6.csv.gz, │ │ Tp.csv.gz, V.csv.gz, Vx.csv.gz, Xl.csv.gz, Xs.csv.gz │ ├── masked-ts-diffusion/ # Causal-attention DDPM implementation used for │ ├── model.py # generative over-sampling │ ├── api.py │ └── __init__.py │ ├── timegan.py # TimeGAN implementation used for generative over-sampling ├── discriminative_metrics.py # Discriminative-score utilities for synthetic-sample evaluation ├── utils.py # Shared helper functions │ ├── 0_Dependancies.ipynb # Environment setup / package installation ├── 1_DataPreparation.ipynb # Builds the MVTS dataset, chronological train/val/test split ├── 2_ImpactsOfNormalization.ipynb # Normalization experiments (Hybrid, Z-score, Min-Max, Log) ├── 2.1_InitialClassification.ipynb # Baseline classifier hyperparameter search ├── 3_BorderlineCleaning.ipynb # Borderline-cleaning experiments (Tomek Links, ENN, NearMiss-3) ├── 4_RandomUnderSampling.ipynb # RUS-alone experiments at several retained-NSEP levels ├── 5_RUS&OS-SMOTE.ipynb # RUS + SMOTE over-sampling experiments ├── 6_RUS&OS-TimeGAN.ipynb # RUS + TimeGAN generation and training ├── 7_RUS&OS-TimeGAN-Results.ipynb # RUS + TimeGAN classification results ├── 8_RUS&OS-Adasyn.ipynb # RUS + ADASYN over-sampling experiments ├── 9_RUS&OS-Diffusion.ipynb # RUS + diffusion-model generation and training ├── 10_RUS&OS-Diffusion-Results.ipynb # RUS + diffusion-model classification results │ ├── 11_Figures_Style.ipynb # Shared plotting style/colors and result-loading utilities ├── 12_Figures_Preprocessing.ipynb # Figure: normalization / cleaning / RUS sweep ├── 13_Figures_DesignChoices.ipynb # Figure: augmentation algorithm comparison ├── 14_Figures_Progression.ipynb # Figure: end-to-end pipeline stage progression ├── 15_Figures_NewInsights.ipynb # Additional analysis figures ├── 16_Figures_Dataset.ipynb # Figure: dataset/class-imbalance overview ├── 17_Figures_Pipeline.ipynb # Figure: pipeline schematic diagram ├── 18_Figures_AugmentationAndBaselines.ipynb # Figure: augmentation trends and published-baseline comparison ├── 19_Figures_TSNE_KDE.ipynb # Figure: t-SNE/KDE comparison of real vs. synthetic SEP sequences │ ├── figures/ # Rendered output figures (see below) ├── requirements / 0_Dependancies.ipynb ├── LICENSE └── README.md