• 제목/요약/키워드: forecast performance

검색결과 521건 처리시간 0.024초

Predicting the CPT-based pile set-up parameters using HHO-RF and PSO-RF hybrid models

  • Yun Dawei;Zheng Bing;Gu Bingbing;Gao Xibo;Behnaz Razzaghzadeh
    • Structural Engineering and Mechanics
    • /
    • 제86권5호
    • /
    • pp.673-686
    • /
    • 2023
  • Determining the properties of pile from cone penetration test (CPT) is costly, and need several in-situ tests. At the present study, two novel hybrid learning models, namely PSO-RF and HHO-RF, which are an amalgamation of random forest (RF) with particle swarm optimization (PSO) and Harris hawks optimization (HHO) were developed and applied to predict the pile set-up parameter "A" from CPT for the design aim of the projects. To forecast the "A," CPT data along were collected from different sites in Louisiana, where the selected variables as input were plasticity index (PI), undrained shear strength (Su), and over consolidation ratio (OCR). Results show that both PSO-RF and HHO-RF models have acceptable performance in predicting the set-up parameter "A," with R2 larger than 0.9094, representing the admissible correlation between observed and predicted values. HHO-RF has better proficiency than the PSO-RF model, with R2 and RMSE equal to 0.9328 and 0.0292 for the training phase and 0.9729 and 0.024 for testing data, respectively. Moreover, PI and OBJ indices are considered, in which the HHO-RF model has lower results which leads to outperforming this hybrid algorithm with respect to PSO-RF for predicting the pile set-up parameter "A," consequently being specified as the proposed model. Therefore, the results demonstrate the ability of the HHO algorithm in determining the optimal value of RF hyperparameters than PSO.

Research on Selecting Influential Climatic Factors and Optimal Timing Exploration for a Rice Production Forecast Model Using Weather Data

  • Jin-Kyeong Seo;Da-Jeong Choi;Juryon Paik
    • 한국컴퓨터정보학회논문지
    • /
    • 제28권7호
    • /
    • pp.57-65
    • /
    • 2023
  • 쌀 생산량 예측의 정확성을 높이기 위한 대다수의 연구는 모델의 정확도 증진에 초점이 맞춰져 있다. 이에 비해, 예측 모델을 적용할 대상 데이터 자체에 관한 연구는 상대적으로 미흡하다. 쌀 생산량 데이터에 동일한 종속변수와 예측 모델을 사용하여 다른 특성들로 구성된 두 부류의 데이터에 적용하면, 결과의 차이가 발생하는데 이때 어느 데이터 셋이 더 우수한지 판단하기는 어려운 일이다. 이러한 문제를 해결하기 위해, 예측 모델 적용 전에 데이터 내에서 예측 결과에 큰 영향을 미칠 가능성이 있는 특성들을 선별하고, 이를 중심으로 모델링을 수행하면, 데이터의 구성이 다르더라도 안정적인 예측 결과를 얻을 수 있을 것이다. 본 연구에서는 기상청의 종관기상관측(ASOS) 데이터를 활용하여, 쌀 생산량의 안정적이고 일관된 예측을 위해 데이터 구성 특성들의 조정을 통해 최적의 기반 변수를 선별하는 방법에 대해 제안한다. 본 연구의 결과는 향후 다른 연구에서 성능평가의 유용성을 높이는 데 기여할 것으로 기대한다.

의약품 콜드체인 유통 수요 예측을 위한 AI 모델에 관한 연구 (A Study on the AI Model for Prediction of Demand for Cold Chain Distribution of Drugs)

  • 김희영;류기환;근재;손현곤
    • 문화기술의 융합
    • /
    • 제9권3호
    • /
    • pp.763-768
    • /
    • 2023
  • 본 논문에서는 의약품 유통량 예측을 위해 기존의 통계 방식(ARIMA)과 머신러닝 방식(Informer)을 개발하고 비교하였다. 일별 데이터의 예측에서는 머신러닝 기반의 모델이 유리하며, 월별 예측에서는 ARIMA를 활용하고 데이터가 증가하면서 Informer로 전환하는 것이 효과적임을 발견하였다. 예측 에러율(RMSE)은 기존 방식 대비 26.6% 낮아졌으며, 예측 정확도도 13% 개선되어 86.2%의 결과를 보였다. 본 논문을 통해 통계적 방법과 머신러닝 방법을 앙상블하여 최상의 결과를 얻을 수 있다는 장점을 발견하였다. 또한 머신러닝 기반의 AI 모델은 불규칙한 상황에서도 딥러닝 연산을 통해 최선의 결과를 도출할 수 있으며, 상용화 이후에는 데이터양이 증가함에 따라 성능이 향상될 것으로 기대된다.

동적요인모형에 기반한 한국의 GDP 성장률 예측 (Forecasting Korea's GDP growth rate based on the dynamic factor model)

  • 이경서;임예지
    • 응용통계연구
    • /
    • 제37권2호
    • /
    • pp.255-263
    • /
    • 2024
  • GDP는 한 나라의 가계, 기업, 정부 등 모든 경제 주체가 일정 기간 동안 창출한 재화와 서비스의 시장 가치의 합을 나타낸다. GDP를 통하여 국가의 경제 규모를 파악할 수 있으며, 정부의 정책 방향에 영향을 미치는 대표적인 경제 지표이므로 이에 대한 연구가 다양하게 이루어지고 있다. 본 논문에서는 G20 국가들의 주요 거시경제 지표를 활용하여 dynamic factor model 기반의 GDP 성장률 예측 모델을 제시하였다. 추출된 factor를 다양한 회귀분석 방법론과 결합하여 그 결과들을 비교하였으며, 기존의 전통적인 시계열 예측방법인 ARIMA 모델, common component를 이용한 예측 등도 함께 비교하였다. COVID 이후 지표의 변동성이 큰 점을 고려하여 예측 시기를 COVID 전후로 나누었으며, 그 결과 factor에 대해 ridge regression과 lasso regression을 적용하여 예측한 경우 가장 좋은 성능을 나타내었다.

Prophet와 GRU을 이용하여 단중기 전력소비량 예측 (Short-and Mid-term Power Consumption Forecasting using Prophet and GRU)

  • 손남례;강은주
    • 스마트미디어저널
    • /
    • 제12권11호
    • /
    • pp.18-26
    • /
    • 2023
  • 빌딩에너지관리시스템(BEMS: Building Energy Management System)은 생산 및 소비되는 에너지를 효율적으로 관리하는 시스템이다. 그러나 건물 내 전력소비는 물리적인 특성상으로 인해 생산 및 소비가 일정하지 않아 안정적인 전력 공급이 필수적이다. 이에 따라 건물의 안정적인 전력 공급을 위해서는 정확한 건물 내 전력 소비 예측이 중요하다. 최근에는 시계열분석, 통계분석, 인공지능 등 다양한 방법을 이용하여 전력소비예측에 관한 연구가 진행되고 있다. 본 논문은 Prophet 모델의 장점과 단점을 분석하여 장점인 growth, seasonality, holidays를 선택하였고, Prophet 모델의 단점인 데이터의 복잡성과 외부변수(기후 데이터)의 제한성을 해결하기 위하여 GRU을 조합하여 단기(2일) 및 중기(7일, 15일, 30일) 전력소비량 예측 알고리즘을 제안한다. 실험결과, 제안한 방법은 기존 GRU 및 Prophet 모델보다 성능이 우수하였다.

In-depth exploration of machine learning algorithms for predicting sidewall displacement in underground caverns

  • Hanan Samadi;Abed Alanazi;Sabih Hashim Muhodir;Shtwai Alsubai;Abdullah Alqahtani;Mehrez Marzougui
    • Geomechanics and Engineering
    • /
    • 제37권4호
    • /
    • pp.307-321
    • /
    • 2024
  • This paper delves into the critical assessment of predicting sidewall displacement in underground caverns through the application of nine distinct machine learning techniques. The accurate prediction of sidewall displacement is essential for ensuring the structural safety and stability of underground caverns, which are prone to various geological challenges. The dataset utilized in this study comprises a total of 310 data points, each containing 13 relevant parameters extracted from 10 underground cavern projects located in Iran and other regions. To facilitate a comprehensive evaluation, the dataset is evenly divided into training and testing subset. The study employs a diverse array of machine learning models, including recurrent neural network, back-propagation neural network, K-nearest neighbors, normalized and ordinary radial basis function, support vector machine, weight estimation, feed-forward stepwise regression, and fuzzy inference system. These models are leveraged to develop predictive models that can accurately forecast sidewall displacement in underground caverns. The training phase involves utilizing 80% of the dataset (248 data points) to train the models, while the remaining 20% (62 data points) are used for testing and validation purposes. The findings of the study highlight the back-propagation neural network (BPNN) model as the most effective in providing accurate predictions. The BPNN model demonstrates a remarkably high correlation coefficient (R2 = 0.99) and a low error rate (RMSE = 4.27E-05), indicating its superior performance in predicting sidewall displacement in underground caverns. This research contributes valuable insights into the application of machine learning techniques for enhancing the safety and stability of underground structures.

Creation of regression analysis for estimation of carbon fiber reinforced polymer-steel bond strength

  • Xiaomei Sun;Xiaolei Dong;Weiling Teng;Lili Wang;Ebrahim Hassankhani
    • Steel and Composite Structures
    • /
    • 제51권5호
    • /
    • pp.509-527
    • /
    • 2024
  • Bonding carbon fiber-reinforced polymer (CFRP) laminates have been extensively employed in the restoration of steel constructions. In addition to the mechanical properties of the CFRP, the bond strength (PU) between the CFRP and steel is often important in the eventual strengthened performance. Nonetheless, the bond behavior of the CFRP-steel (CS) interface is exceedingly complicated, with multiple failure causes, giving the PU challenging to forecast, and the CFRP-enhanced steel structure is unsteady. In just this case, appropriate methods were established by hybridized Random Forests (RF) and support vector regression (SVR) approaches on assembled CS single-shear experiment data to foresee the PU of CS, in which a recently established optimization algorithm named Aquila optimizer (AO) was used to tune the RF and SVR hyperparameters. In summary, the practical novelty of the article lies in its development of a reliable and efficient method for predicting bond strength at the CS interface, which has significant implications for structural rehabilitation, design optimization, risk mitigation, cost savings, and decision support in engineering practice. Moreover, the Fourier Amplitude Sensitivity Test was performed to depict each parameter's impact on the target. The order of parameter importance was tc> Lc > EA > tA > Ec > bc > fc > fA from largest to smallest by 0.9345 > 0.8562 > 0.79354 > 0.7289 > 0.6531 > 0.5718 > 0.4307 > 0.3657. In three training, testing, and all data phases, the superiority of AO - RF with respect to AO - SVR and MARS was obvious. In the training stage, the values of R2 and VAF were slightly similar with a tiny superiority of AO - RF compared to AO - SVR with R2 equal to 0.9977 and VAF equal to 99.772, but large differences with results of MARS.

Large-scale Atmospheric Patterns associated with the 2018 Heatwave Prediction in the Korea-Japan Region using GloSea6

  • Jinhee Kang;Semin Yun;Jieun Wie;Sang-Min Lee;Johan Lee;Baek-Jo Kim;Byung-Kwon Moon
    • 한국지구과학회지
    • /
    • 제45권1호
    • /
    • pp.37-47
    • /
    • 2024
  • In the summer of 2018, the Korea-Japan (KJ) region experienced an extremely severe and prolonged heatwave. This study examines the GloSea6 model's prediction performance for the 2018 KJ heatwave event and investigates how its prediction skill is related to large-scale circulation patterns identified by the k-means clustering method. Cluster 1 pattern is characterized by a KJ high-pressure anomaly, Cluster 2 pattern is distinguished by an Eastern European high-pressure anomaly, and Cluster 3 pattern is associated with a Pacific-Japan pattern-like anomaly. By analyzing the spatial correlation coefficients between these three identified circulation patterns and GloSea6 predictions, we assessed the contribution of each circulation pattern to the heatwave lifecycle. Our results show that the Eastern European high-pressure pattern, in particular, plays a significant role in predicting the evolution of the development and peak phases of the 2018 KJ heatwave approximately two weeks in advance. Furthermore, this study suggests that an accurate representation of large-scale atmospheric circulations in upstream regions is a key factor in seasonal forecast models for improving the predictability of extreme weather events, such as the 2018 KJ heatwave.

Prediction of rock slope failure using multiple ML algorithms

  • Bowen Liu;Zhenwei Wang;Sabih Hashim Muhodir;Abed Alanazi;Shtwai Alsubai;Abdullah Alqahtani
    • Geomechanics and Engineering
    • /
    • 제36권5호
    • /
    • pp.489-509
    • /
    • 2024
  • Slope stability analysis and prediction are of critical importance to geotechnical engineers, given the severe consequences associated with slope failure. This research endeavors to forecast the factor of safety (FOS) for slopes through the implementation of six distinct ML techniques, including back propagation neural networks (BPNN), feed-forward neural networks (FFNN), Takagi-Sugeno fuzzy system (TSF), gene expression programming (GEP), and least-square support vector machine (Ls-SVM). 344 slope cases were analyzed, incorporating a variety of geometric and shear strength parameters measured through the PLAXIS software alongside several loss functions to assess the models' performance. The findings demonstrated that all models produced satisfactory results, with BPNN and GEP models proving to be the most precise, achieving an R2 of 0.86 each and MAE and MAPE rates of 0.00012 and 0.00002 and 0.005 and 0.004, respectively. A Pearson correlation and residuals statistical analysis were carried out to examine the importance of each factor in the prediction, revealing that all considered geomechanical features are significantly relevant to slope stability. However, the parameters of friction angle and slope height were found to be the most and least significant, respectively. In addition, to aid in the FOS computation for engineering challenges, a graphical user interface (GUI) for the ML-based techniques was created.

Application of data-driven model reduction techniques in reactor neutron field calculations

  • Zhaocai Xiang;Qiafeng Chen;Pengcheng Zhao
    • Nuclear Engineering and Technology
    • /
    • 제56권8호
    • /
    • pp.2948-2957
    • /
    • 2024
  • High-order harmonic techniques can be used to recreate neutron flux distributions in reactor cores using the neutron diffusion equation. However, traditional source iteration and source correction iteration techniques have sluggish convergence rates and protracted calculation periods. The correctness of the implicitly restarted Arnoldi method (IRAM) in resolving the eigenvalue problems of the one-dimensional and two-dimensional neutron diffusion equations was confirmed by computing the benchmark problems SLAB_1D_1G and two-dimensional steady-state TWIGL using IRAM. By integrating Galerkin projection with Proper Orthogonal Decomposition (POD) techniques, a POD-Galerkin reduced-order model was developed and the IRAM model was used as the full-order model. For 14 macroscopic cross-section values, the TWIGL benchmark problem was perturbed within a 20% range. We extracted 100 sample points using the Latin hypercube sampling method, and 70% of the samples were used as the testing set to assess the performance of the reduced-order model The remaining 30% were utilized as the training set to develop the reduced-order model, which was employed to rebuild the TWIGL benchmark problem. The reduced-order model demonstrates good flexibility and can efficiently and accurately forecast the effective multiplication factor and neutron flux distribution in the core. The reduced-order model predicts keff and neutron flux distribution with a high degree of agreement compared to the full-order model. Additionally, the reduced-order model's computation time is only 10.18% of that required by the full-order model.The neutron flux distribution of the steady-state TWIGL benchmark was recreated using the reduced-order model. The obtained results indicate that the reduced-order model can accurately predict the keff and neutron flux distribution of the steady-state TWIGL benchmark.Overall, the proposed technique not only has the potential to accurately project neutron flux distributions in transient settings, but is also relevant for reconstructing neutron flux distributions in steady-state conditions; thus, its applicability is bound to increase in the future.