Parallel boosting neural network with mutual information for day-ahead solar irradiance forecasting
Bibliographic record
Abstract
The transition to sustainable energy has become imperative due to the depletion of fossil fuels. Solar energy presents a viable alternative owing to its abundance and environmental benefits. However, the intermittent nature of solar energy requires accurate forecasting of solar irradiance (SI) for reliable operation of photovoltaics (PVs) integrated systems. Traditional deep learning (DL) models and decision tree (DT)-based algorithms have been widely employed for this purpose. However, DL models often demand substantial computational resources and large datasets, while DT algorithms lack generalizability. To address these limitations, this study proposes a novel parallel boosting neural network (PBNN) framework that integrates boosting algorithms with a feedforward neural network (FFNN). The proposed framework leverages three boosting DT algorithms, Extreme Gradient Boosting (XgBoost), Categorical Boosting (CatBoost), and Random Forest (RF) regressors as base learners, operating in parallel. The intermediary forecasts from these base learners are concatenated and input into the FFNN, which assigns optimal weights to generate the final prediction. The proposed PBNN is trained and evaluated on two geographical datasets and compared with state-of-the-art techniques. The mutual information (MI) algorithm is implemented as a feature selection technique to identify the most important features for forecasting. Results demonstrate that when trained with the selected features, the mean absolute percentage error (MAPE) of PBNN is improved by [Formula: see text], and [Formula: see text] for Islamabad and San Diego city datasets, respectively. Furthermore, a literature comparison of the PBNN is also performed for robustness analysis. Source code and datasets are available at https://github.com/Ubaid014/Parallel-Boosting-Neural-Network/tree/main.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".