pith. sign in

arxiv: 2408.00601 · v1 · submitted 2024-08-01 · 💻 cs.LG

AutoPV: Automatically Design Your Photovoltaic Power Forecasting Model

Dayin Chen , Xiaodan Shi , Mingkun Jiang , Haoran Zhang , Dongxiao Zhang , Yuntian Chen , Jinyue Yan This is my paper

Pith reviewed 2026-05-23 22:00 UTC · model grok-4.3

classification 💻 cs.LG
keywords photovoltaic power forecastingneural architecture searchautomated model designtime series forecastingdeep learning modelssolar energy predictionNAS framework
0
0 comments X

The pith

AutoPV uses neural architecture search to automatically build photovoltaic power forecasting models that outperform state-of-the-art hand-designed ones.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents AutoPV as a framework that applies neural architecture search to create custom models for photovoltaic power forecasting tasks. It builds a new search space by combining data processing methods drawn from leading time series forecasting models and existing solar power deep learning approaches. Tests on data from the Daqing Photovoltaic Station show the resulting architectures deliver better predictions than current top models and complete the design process quickly. This addresses the difficulty of needing cross-domain expertise and manual effort to craft effective forecasting systems for solar energy. A sympathetic reader would care because it shifts model creation from labor-intensive expert work to an automated process that industries could run directly.

Core claim

AutoPV is a NAS-based framework that automates the design of PVPF models by searching within a novel space combining techniques from SOTA TSF and PVPF models. Experiments on the Daqing Photovoltaic Station dataset show it constructs superior architectures in relatively short time compared to predefined SOTA models.

What carries the argument

The novel NAS search space that incorporates various data processing techniques from state-of-the-art TSF models and typical PVPF deep learning models.

If this is right

  • Non-experts can obtain effective PVPF models without requiring cross-domain knowledge.
  • Industries can reduce the labor costs involved in constructing predictive architectures for solar power tasks.
  • The approach bridges the application of NAS techniques to time series forecasting problems.
  • Automated construction produces models better matched to specific PVPF tasks than manually predefined ones.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same search-space construction could be tested on related renewable forecasting problems such as wind power output.
  • If the space proves reusable, similar NAS setups might speed iteration in other energy-related time series applications.
  • Wider use could shift practice away from fixed benchmark models toward task-specific automated designs.

Load-bearing premise

The newly designed search space is broad and well-structured enough to contain architectures meaningfully better than existing hand-crafted ones rather than merely rediscovering variants reachable by manual tuning.

What would settle it

Running AutoPV on the Daqing Photovoltaic Station dataset yields models that fail to outperform the SOTA predefined models or requires substantially more time than reported.

Figures

Figures reproduced from arXiv: 2408.00601 by Dayin Chen, Dongxiao Zhang, Haoran Zhang, Jinyue Yan, Mingkun Jiang, Xiaodan Shi, Yuntian Chen.

Figure 1
Figure 1. Figure 1: The pipeline for leveraging AutoPV in PVPF tasks involves two distinct scenarios: PVPF Task 1, which utilizes only historical data series as input, [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A minor modification in the core predictive structure of AutoPV. In typical TSF models, the forecasting results for all features are generated first, then [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The detailed process of neural architecture search and construction. The FFM and FFT parameters determine the final feature set. The DGM, SM, and [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Search performance of different PVPF tasks. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The average search time of each iteration and the search wall time [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: The complete architecture searched by AutoPV for the different sub-tasks in PVPF Task 1. [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The complete architecture searched by AutoPV for the different tasks in PVPF Task 2. [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visualization of AutoPV forecasting performance: with and without future weather data [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 1
Figure 1. Figure 1: The optimal architectures found during each iteration for PVPF Task 1. They constitute the Pareto Frontier. The yellow points on the figures indicate [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The optimal architectures found during each iteration for PVPF Task 2. They constitute the Pareto Frontier. The yellow points on the figures indicate [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Proportions of the 12 parameters in the AutoPV search space. The statistics encompass the final optimal set of the searched neural architectures across [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
read the original abstract

Photovoltaic power forecasting (PVPF) is a critical area in time series forecasting (TSF), enabling the efficient utilization of solar energy. With advancements in machine learning and deep learning, various models have been applied to PVPF tasks. However, constructing an optimal predictive architecture for specific PVPF tasks remains challenging, as it requires cross-domain knowledge and significant labor costs. To address this challenge, we introduce AutoPV, a novel framework for the automated search and construction of PVPF models based on neural architecture search (NAS) technology. We develop a brand new NAS search space that incorporates various data processing techniques from state-of-the-art (SOTA) TSF models and typical PVPF deep learning models. The effectiveness of AutoPV is evaluated on diverse PVPF tasks using a dataset from the Daqing Photovoltaic Station in China. Experimental results demonstrate that AutoPV can complete the predictive architecture construction process in a relatively short time, and the newly constructed architecture is superior to SOTA predefined models. This work bridges the gap in applying NAS to TSF problems, assisting non-experts and industries in automatically designing effective PVPF models.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes AutoPV, a neural architecture search (NAS) framework for automatically constructing photovoltaic power forecasting (PVPF) models. It defines a new search space that incorporates data-processing and architectural components from state-of-the-art time-series forecasting (TSF) and PVPF models, then evaluates the resulting architectures on a dataset from the Daqing Photovoltaic Station, claiming that the NAS process completes quickly and yields models superior to hand-crafted SOTA baselines.

Significance. If the empirical superiority claim is substantiated with rigorous validation, the work would demonstrate a practical application of NAS to PVPF, potentially lowering the barrier for non-experts to obtain effective forecasting models. The contribution lies in the empirical demonstration of automated design rather than a new theoretical NAS method; credit is due for targeting a real-world energy application, though the absence of reproducible experimental details limits immediate impact.

major comments (2)
  1. [Abstract and Experimental Results] Abstract and Experimental Results section: The central claim that the NAS-derived architecture is superior to SOTA predefined models is presented without any reported metrics (e.g., RMSE, MAE, or MAPE), baseline model names, cross-validation procedure, number of independent NAS runs with different seeds, or statistical significance tests. This information is load-bearing for evaluating whether the performance gain is reliable or reproducible.
  2. [Search-space construction] Search-space construction (likely §3): The space is explicitly assembled from techniques already present in SOTA TSF and PVPF models. Without an ablation or comparison showing that the discovered architecture could not have been reached by targeted manual combination and tuning of the same components, it remains unclear whether the reported gains demonstrate the value of NAS or simply the value of the expanded component library itself.
minor comments (2)
  1. [Abstract] The abstract states evaluation on 'diverse PVPF tasks' yet references only a single station dataset; clarify whether additional datasets or task variations were used and report results accordingly.
  2. [Method] Notation for the search space components and the NAS controller (e.g., any reward function or sampling strategy) should be defined explicitly with equations or pseudocode to allow replication.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight important aspects for strengthening the presentation of our empirical results and clarifying the contribution of the NAS framework. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract and Experimental Results] Abstract and Experimental Results section: The central claim that the NAS-derived architecture is superior to SOTA predefined models is presented without any reported metrics (e.g., RMSE, MAE, or MAPE), baseline model names, cross-validation procedure, number of independent NAS runs with different seeds, or statistical significance tests. This information is load-bearing for evaluating whether the performance gain is reliable or reproducible.

    Authors: We agree that additional details are necessary to fully substantiate the superiority claims. Although the Experimental Results section includes comparative evaluations on the Daqing dataset, we will revise the manuscript to explicitly report the numerical metrics (RMSE, MAE, MAPE), list all baseline model names, describe the cross-validation procedure, specify the number of independent NAS runs with different seeds, and include statistical significance tests (e.g., paired t-tests or Wilcoxon tests). These changes will improve reproducibility and allow readers to assess the reliability of the gains. revision: yes

  2. Referee: [Search-space construction] Search-space construction (likely §3): The space is explicitly assembled from techniques already present in SOTA TSF and PVPF models. Without an ablation or comparison showing that the discovered architecture could not have been reached by targeted manual combination and tuning of the same components, it remains unclear whether the reported gains demonstrate the value of NAS or simply the value of the expanded component library itself.

    Authors: The search space is deliberately constructed from components in existing SOTA models, as described in §3, to incorporate proven techniques from both TSF and PVPF literature. The core contribution of AutoPV is the NAS-driven automation that enables non-experts to efficiently identify high-performing architectures without labor-intensive manual tuning and cross-domain expertise. The reported results show that the discovered architecture outperforms the hand-crafted SOTA baselines, illustrating the practical benefit of systematic search. We maintain that the gains reflect the value of the automated process rather than solely the component library, as manual exploration of the same space would be time-consuming and expertise-dependent; the manuscript already frames the work as an application of NAS to PVPF rather than a novel NAS algorithm. revision: no

Circularity Check

0 steps flagged

No circularity: empirical NAS framework with no derivation chain

full rationale

The paper describes an empirical application of neural architecture search to photovoltaic power forecasting. It defines a search space by combining existing data-processing and architectural techniques from prior TSF and PVPF literature, then reports experimental performance of the discovered architectures versus hand-crafted SOTA models. No equations, first-principles derivations, fitted parameters renamed as predictions, or uniqueness theorems appear. The central claim rests on experimental comparisons rather than any reduction to self-defined inputs or self-citation chains. The search-space construction is explicit and external to the performance metric, so the reported superiority does not reduce by construction to the authors' own definitions.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no concrete equations, model specifications, or experimental details from which to extract free parameters, axioms, or invented entities. The central claim rests on the unstated assumption that NAS search will discover superior models within the proposed space.

pith-pipeline@v0.9.0 · 5745 in / 1114 out tokens · 29525 ms · 2026-05-23T22:00:42.414951+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

41 extracted references · 41 canonical work pages · 5 internal anchors

  1. [1]

    Day-ahead probabilistic forecasting at a co-located wind and solar power park in sweden: Trading and forecast verification,

    O. Lindberg, D. Lingfors, J. Arnqvist, D. van Der Meer, and J. Munkhammar, “Day-ahead probabilistic forecasting at a co-located wind and solar power park in sweden: Trading and forecast verification,” Advances in Applied Energy , vol. 9, p. 100120, 2023

    work page 2023

  2. [2]

    Evaluating neural network models in site-specific solar pv forecasting using numerical weather prediction data and weather observations,

    C. Brester, V . Kallio-Myers, A. V . Lindfors, M. Kolehmainen, and H. Niska, “Evaluating neural network models in site-specific solar pv forecasting using numerical weather prediction data and weather observations,” Renewable Energy, vol. 207, pp. 266–274, 2023

    work page 2023

  3. [3]

    Deep learning neural networks for short-term photovoltaic power forecasting,

    A. Mellit, A. M. Pavan, and V . Lughi, “Deep learning neural networks for short-term photovoltaic power forecasting,” Renewable Energy, vol. 172, pp. 276–288, 2021

    work page 2021

  4. [4]

    Day-ahead power forecasting in a large-scale photovoltaic plant based on weather classification using lstm,

    M. Gao, J. Li, F. Hong, and D. Long, “Day-ahead power forecasting in a large-scale photovoltaic plant based on weather classification using lstm,” Energy, vol. 187, p. 115838, 2019

    work page 2019

  5. [5]

    A tcn-based hybrid forecasting framework for hours- ahead utility-scale pv forecasting,

    Y . Li, L. Song, S. Zhang, L. Kraus, T. Adcox, R. Willardson, A. Koman- dur, and N. Lu, “A tcn-based hybrid forecasting framework for hours- ahead utility-scale pv forecasting,” IEEE Transactions on Smart Grid , 2023

    work page 2023

  6. [6]

    The rise of data-driven weather forecasting: A first statistical assessment of machine learning-based weather forecasts in an operational-like context,

    Z. Ben Bouall `egue, M. C. Clare, L. Magnusson, E. Gascon, M. Maier- Gerber, M. Janou ˇsek, M. Rodwell, F. Pinault, J. S. Dramsch, S. T. Lang et al. , “The rise of data-driven weather forecasting: A first statistical assessment of machine learning-based weather forecasts in an operational-like context,” Bulletin of the American Meteorological Society, 2024

    work page 2024

  7. [7]

    Skygpt: Probabilistic ultra-short-term solar forecasting using synthetic sky im- ages from physics-constrained videogpt,

    Y . Nie, E. Zelikman, A. Scott, Q. Paletta, and A. Brandt, “Skygpt: Probabilistic ultra-short-term solar forecasting using synthetic sky im- ages from physics-constrained videogpt,” Advances in Applied Energy , vol. 14, p. 100172, 2024

    work page 2024

  8. [8]

    A review and evaluation of the state-of-the-art in pv solar power forecasting: Techniques and optimization,

    R. Ahmed, V . Sreeram, Y . Mishra, and M. Arif, “A review and evaluation of the state-of-the-art in pv solar power forecasting: Techniques and optimization,” Renewable and Sustainable Energy Reviews , vol. 124, p. 109792, 2020

    work page 2020

  9. [9]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems , vol. 30, 2017

    work page 2017

  10. [10]

    Informer: Beyond efficient transformer for long sequence time-series forecasting,

    H. Zhou, S. Zhang, J. Peng, S. Zhang, J. Li, H. Xiong, and W. Zhang, “Informer: Beyond efficient transformer for long sequence time-series forecasting,” in Proceedings of the AAAI conference on artificial intel- ligence, vol. 35, no. 12, 2021, pp. 11 106–11 115

    work page 2021

  11. [11]

    Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting,

    H. Wu, J. Xu, J. Wang, and M. Long, “Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting,” Advances in neural information processing systems, vol. 34, pp. 22 419– 22 430, 2021

    work page 2021

  12. [12]

    Fedformer: Frequency enhanced decomposed transformer for long-term series fore- casting,

    T. Zhou, Z. Ma, Q. Wen, X. Wang, L. Sun, and R. Jin, “Fedformer: Frequency enhanced decomposed transformer for long-term series fore- casting,” in International conference on machine learning . PMLR, 2022, pp. 27 268–27 286

    work page 2022

  13. [13]

    Are transformers effective for time series forecasting?

    A. Zeng, M. Chen, L. Zhang, and Q. Xu, “Are transformers effective for time series forecasting?” in Proceedings of the AAAI conference on artificial intelligence, vol. 37, no. 9, 2023, pp. 11 121–11 128

    work page 2023

  14. [14]

    Tsmixer: An all-mlp architecture for time series forecast-ing,

    S.-A. Chen, C.-L. Li, S. O. Arik, N. C. Yoder, and T. Pfister, “Tsmixer: An all-mlp architecture for time series forecast-ing,” Transactions on Machine Learning Research , 2023

    work page 2023

  15. [15]

    Neural Architecture Search with Reinforcement Learning

    B. Zoph and Q. V . Le, “Neural architecture search with reinforcement learning,” arXiv preprint arXiv:1611.01578 , 2016

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  16. [16]

    Client: Cross-variable linear integrated enhanced transformer for multivariate long-term time series forecasting,

    J. Gao, W. Hu, and Y . Chen, “Client: Cross-variable linear integrated enhanced transformer for multivariate long-term time series forecasting,” 2023

    work page 2023

  17. [17]

    Frequency-domain mlps are more effective learners in time series forecasting,

    K. Yi, Q. Zhang, W. Fan, S. Wang, P. Wang, H. He, N. An, D. Lian, L. Cao, and Z. Niu, “Frequency-domain mlps are more effective learners in time series forecasting,” Advances in Neural Information Processing Systems, vol. 36, 2024

    work page 2024

  18. [18]

    ETSformer: Exponential Smoothing Transformers for Time-series Forecasting,

    G. Woo, C. Liu, D. Sahoo, A. Kumar, and S. Hoi, “Etsformer: Exponen- tial smoothing transformers for time-series forecasting,” arXiv preprint arXiv:2202.01381, 2022

    work page arXiv 2022

  19. [19]

    A Time Series is Worth 64 Words: Long-term Forecasting with Transformers

    Y . Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam, “A time series is worth 64 words: Long-term forecasting with transformers,” arXiv preprint arXiv:2211.14730, 2022

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  20. [20]

    iTransformer: Inverted Transformers Are Effective for Time Series Forecasting

    Y . Liu, T. Hu, H. Zhang, H. Wu, S. Wang, L. Ma, and M. Long, “itrans- former: Inverted transformers are effective for time series forecasting,” arXiv preprint arXiv:2310.06625 , 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  21. [21]

    Micn: Multi-scale local and global context modeling for long-term series forecasting,

    H. Wang, J. Peng, F. Huang, J. Wang, J. Chen, and Y . Xiao, “Micn: Multi-scale local and global context modeling for long-term series forecasting,” in The Eleventh International Conference on Learning Representations, 2022

    work page 2022

  22. [22]

    Timesnet: Temporal 2d-variation modeling for general time series analysis,

    H. Wu, T. Hu, Y . Liu, H. Zhou, J. Wang, and M. Long, “Timesnet: Temporal 2d-variation modeling for general time series analysis,” in The eleventh international conference on learning representations , 2022

    work page 2022

  23. [23]

    A survey on diffusion models for time series and spatio-temporal data,

    Y . Yang, M. Jin, H. Wen, C. Zhang, Y . Liang, L. Ma, Y . Wang, C. Liu, B. Yang, Z. Xu et al. , “A survey on diffusion models for time series and spatio-temporal data,” arXiv preprint arXiv:2404.18886 , 2024

    work page arXiv 2024

  24. [24]

    Efficient neural architecture search via parameters sharing,

    H. Pham, M. Guan, B. Zoph, Q. Le, and J. Dean, “Efficient neural architecture search via parameters sharing,” in International conference on machine learning . PMLR, 2018, pp. 4095–4104

    work page 2018

  25. [25]

    Deeper insights into weight sharing in neural architecture search,

    Y . Zhang, Z. Lin, J. Jiang, Q. Zhang, Y . Wang, H. Xue, C. Zhang, and Y . Yang, “Deeper insights into weight sharing in neural architecture search,” arXiv preprint arXiv:2001.01431 , 2020

    work page arXiv 2001

  26. [26]

    Aging evolution for image classifier architecture search,

    E. Real, A. Aggarwal, Y . Huang, and Q. V . Le, “Aging evolution for image classifier architecture search,” in AAAI conference on artificial intelligence, vol. 2, 2019, p. 2

    work page 2019

  27. [27]

    Bananas: Bayesian opti- mization with neural architectures for neural architecture search,

    C. White, W. Neiswanger, and Y . Savani, “Bananas: Bayesian opti- mization with neural architectures for neural architecture search,” in 12 0 20 40 60 80 100 Time (h) 0 200 400 600 800 1000Power Generation(kW) 2022-08-01 00:00 Prediction without Future Weather Prediction with Future Weather Ground Truth (a) DQH-12 0 20 40 60 80 100 120 Time (h) 0 200 400 ...

    work page 2022

  28. [28]

    Bag of baselines for multi-objective joint neural architecture search and hyperparameter optimization,

    S. Izquierdo, J. Guerrero-Viu, S. Hauns, G. Miotto, S. Schrodi, A. Biedenkapp, T. Elsken, D. Deng, M. Lindauer, and F. Hutter, “Bag of baselines for multi-objective joint neural architecture search and hyperparameter optimization,” in 8th ICML Workshop on Automated Machine Learning (AutoML) , 2021

    work page 2021

  29. [29]

    Atpfl: Automatic trajectory prediction model design under federated learning framework,

    C. Wang, X. Chen, J. Wang, and H. Wang, “Atpfl: Automatic trajectory prediction model design under federated learning framework,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6563–6572

    work page 2022

  30. [30]

    Feath- ers: Federated architecture and hyperparameter search,

    J. Seng, P. Prasad, M. Mundt, D. S. Dhami, and K. Kersting, “Feath- ers: Federated architecture and hyperparameter search,” arXiv preprint arXiv:2206.12342, 2022

    work page arXiv 2022

  31. [31]

    HAT: Hardware-aware transformers for efficient natural language processing,

    H. Wang, Z. Wu, Z. Liu, H. Cai, L. Zhu, C. Gan, and S. Han, “HAT: Hardware-aware transformers for efficient natural language processing,” in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , D. Jurafsky, J. Chai, N. Schluter, and J. Tetreault, Eds. Online: Association for Computational Linguistics, Jul. 2020, pp. 76...

    work page 2020

  32. [32]

    Machine learning for forecasting a photovoltaic (pv) generation system,

    C. Scott, M. Ahsan, and A. Albarbar, “Machine learning for forecasting a photovoltaic (pv) generation system,” Energy, vol. 278, p. 127807, 2023

    work page 2023

  33. [33]

    Prediction of solar energy guided by pearson correlation using machine learning,

    I. Jebli, F.-Z. Belouadha, M. I. Kabbaj, and A. Tilioua, “Prediction of solar energy guided by pearson correlation using machine learning,” Energy, vol. 224, p. 120109, 2021

    work page 2021

  34. [34]

    Reversible instance normalization for accurate time-series forecasting against distribution shift,

    T. Kim, J. Kim, Y . Tae, C. Park, J.-H. Choi, and J. Choo, “Reversible instance normalization for accurate time-series forecasting against distribution shift,” in International Conference on Learning Representations, 2022. [Online]. Available: https://openreview.net/ forum?id=cGDAkQo1C0p

    work page 2022

  35. [35]

    Deep adaptive input normalization for time series forecasting,

    N. Passalis, A. Tefas, J. Kanniainen, M. Gabbouj, and A. Iosifidis, “Deep adaptive input normalization for time series forecasting,” IEEE Transactions on Neural Networks and Learning Systems , vol. 31, no. 9, pp. 3760–3765, 2020

    work page 2020

  36. [36]

    Predicting photovoltaic power production using high-uncertainty weather forecasts,

    T. Polasek and M. Cadk, “Predicting photovoltaic power production using high-uncertainty weather forecasts,” Applied Energy, vol. 339, p. 120989, 2023

    work page 2023

  37. [37]

    ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware

    H. Cai, L. Zhu, and S. Han, “ProxylessNAS: Direct neural architecture search on target task and hardware,” in International Conference on Learning Representations , 2019. [Online]. Available: https://arxiv.org/pdf/1812.00332.pdf

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  38. [38]

    A fast elitist non- dominated sorting genetic algorithm for multi-objective optimization: Nsga-ii,

    K. Deb, S. Agrawal, A. Pratap, and T. Meyarivan, “A fast elitist non- dominated sorting genetic algorithm for multi-objective optimization: Nsga-ii,” in Parallel Problem Solving from Nature PPSN VI: 6th Inter- national Conference Paris, France, September 18–20, 2000 Proceedings

    work page 2000

  39. [39]

    Springer, 2000, pp. 849–858

    work page 2000

  40. [40]

    A tutorial on thompson sampling,

    D. J. Russo, B. Van Roy, A. Kazerouni, I. Osband, Z. Wen et al. , “A tutorial on thompson sampling,” Foundations and Trends® in Machine Learning, vol. 11, no. 1, pp. 1–96, 2018

    work page 2018

  41. [41]

    AutoPV: Automatically Design Your Photovoltaic Power Forecasting Model

    Y . Wang, H. Wu, J. Dong, Y . Liu, M. Long, and J. Wang, “Deep time series models: A comprehensive survey and benchmark,” 2024. 13 VI. B IOGRAPHY SECTION Dayin Chen is currently pursuing his Ph.D. in Building Energy and Environment Engineering at The Hong Kong Polytechnic University. He holds both a Bachelor’s and a Master’s degree in Computer Science fro...

    work page internal anchor Pith review Pith/arXiv arXiv 2024