Performance Assessment of Sequential Models for Solar Radiation Forecasting Over Varying Forecast Horizon

Lecture notes in networks and systems, Год журнала: 2025, Номер unknown, С. 49 - 59

Опубликована: Янв. 1, 2025

Язык: Английский

---

AVI-Net: Audio-visual-integration inspired deep network with application to short-term air temperature forecasting

Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 127604 - 127604

Опубликована: Апрель 1, 2025

Язык: Английский

---

Attention enhanced hybrid model for spatiotemporal short-term forecasting of particulate matter concentrations

Sustainable Cities and Society, Год журнала: 2022, Номер 86, С. 104112 - 104112

Опубликована: Авг. 12, 2022

Язык: Английский

---

Performance Evaluation of Swin Vision Transformer Model Using Gradient Accumulation Optimization Technique

Lecture notes in networks and systems, Год журнала: 2023, Номер unknown, С. 56 - 64

Опубликована: Янв. 1, 2023

Язык: Английский

---

Predictive Analytics of Air Temperature in Alaskan Permafrost Terrain Leveraging Two-Level Signal Decomposition and Deep Learning

Forecasting, Год журнала: 2024, Номер 6(1), С. 55 - 80

Опубликована: Янв. 9, 2024

Local weather forecasts in the Arctic outside of settlements are challenging due to dearth ground-level observation stations and high computational costs. During winter, these critical help prepare for potentially hazardous conditions, while spring, may be used determine flood risk during annual snow melt. To this end, a hybrid VMD-WT-InceptionTime model is proposed multi-horizon multivariate forecasting remote-region temperatures Alaska over short-term horizons (the next seven days). First, Spearman correlation coefficient employed analyze relationship between each input variable forecast target temperature. The most output-correlated sequences decomposed using variational mode decomposition (VMD) and, ultimately, wavelet transform (WT) extract time-frequency patterns intrinsic raw inputs. resulting fed into deep InceptionTime forecasting. This technique has been developed evaluated 35+ years data from three locations Alaska. Different experiments performance benchmarks conducted learning models (e.g., Time Series Transformers, LSTM, MiniRocket), statistical conventional machine baselines GBDT, SVR, ARIMA). All performances assessed four metrics: root mean squared error, absolute percentage determination, directional accuracy. Superior achieved consistently technique.

Язык: Английский

---

Future forecast of global mean surface temperature using machine learning and conventional time series methods

Theoretical and Applied Climatology, Год журнала: 2024, Номер 156(1)

Опубликована: Дек. 12, 2024

Язык: Английский

---

Advancements in weather forecasting for precision agriculture: From statistical modeling to transformer-based architectures

Stochastic Environmental Research and Risk Assessment, Год журнала: 2024, Номер 38(9), С. 3695 - 3717

Опубликована: Авг. 12, 2024

Язык: Английский

---

MTTF: a multimodal transformer for temperature forecasting

International Journal of Computers and Applications, Год журнала: 2023, Номер 46(2), С. 122 - 135

Опубликована: Дек. 8, 2023

Accurate weather forecasting is crucial for various applications, including agriculture and environmental monitoring. However, existing deep learning based methods typically use only temperature observations as input, which do not consider spatial location (e.g. neighboring regions usually show similar trends) increase the difficulty in predicting anomalous fluctuations temperature. To address issue, a multimodal transformer model proposed. This uses observational data (time series data) input but also incorporates spatially rich ERA5 reanalysis (spatio-temporal data). The proposed method has two distinctive features: (1) sequence merging module to highlight dominating features reduce cost of attention calculations. (2) cross-all-modal mechanism capture dependencies between current modality all modalities. was evaluated on three datasets from plain challenging high-altitude mountainous regions. Results showed that outperforms terms MAE MSE, offering promising new solution prediction. Our code are available at https://github.com/Adam618/MTTF.

Язык: Английский

---

An efficient astronomical seeing forecasting method by random convolutional Kernel transformation

Engineering Applications of Artificial Intelligence, Год журнала: 2023, Номер 127, С. 107259 - 107259

Опубликована: Окт. 11, 2023

Язык: Английский

---

Developing seasonal z-number regression for waste-disposal forecasting in a Taiwanese hospital

Applied Soft Computing, Год журнала: 2023, Номер 152, С. 111196 - 111196

Опубликована: Дек. 27, 2023

Язык: Английский

---