Deep Learning vs. Statistical Models for Multi-Horizon Price Forecasting of Second-Hand Electronics: A Systematic Benchmark

By Mateusz Buczyński, Michał Woźniak, Konrad Kaczyński, Anna Wróblewska, Sebastian Kuk

Rating

1753
Battle Count: 50

Relevance

3/10
While the paper focuses on second-hand electronics rather than financial assets, the methodology for multi-horizon forecasting and model benchmarking is transferable. The findings on N-BEATS/N-HiTS superiority over transformers in sparse, volatile markets are relevant for alternative data trading strategies.

Implementation Complexity

6/10
Requires setting up a complex pipeline for web scraping, NLP-based entity matching, and time-series smoothing. Model training involves extensive hyperparameter tuning (Ray Tune) and GPU resources. The evaluation framework with three protocols adds significant complexity.

Reproducibility

2/5
The dataset is proprietary and not publicly available. While the methodology and hyperparameter search ranges are described, the specific data used for training and testing cannot be obtained by external researchers. No public code repository is linked in the text.

Abstract

Forecasting resale prices of used electronics is critical for subscription-based platforms where pricing errors translate directly into risk. Unlike structured financial markets, second-hand electronics exhibit high volatility, sparse listing histories, and non-normal price dynamics - yet no systematic time-series benchmark exists for this domain. This paper presents the first multi-horizon benchmark of statistical and deep learning forecasting models for used electronics price prediction. We use a large-scale dataset of daily price listings from Polish online marketplaces (January 2022 to March 2025, 100+ smartphone and laptop models) and evaluate eleven models across six horizons from 1 to 365 days, covering classical methods (ARIMA, ETS, Theta), recurrent and convolutional networks (LSTM, TCN), and modern deep architectures (N-BEATS, N-HiTS, TFT, PatchTST, Informer). Three complementary evaluation protocols assess trajectory fitness, one-shot endpoint accuracy, and cross-horizon transfer. N-BEATS achieves the lowest MAPE beyond 30 days, reaching 8.51% at 365 days versus 14.94% for the best statistical baseline - a 43% reduction. At short horizons (1-7 days), all models converge near 0.72% MAPE and the naive baseline remains competitive. A single N-BEATS model trained at 365 days generalizes to all shorter horizons, eliminating the need for horizon-specific models. N-BEATS and N-HiTS also demonstrate superior hyperparameter stability.

Links

arXiv:2610.10727 · Paper rankings

The interactive Everscope explorer (charts, battles, favorites) loads below.