SNDT WOMEN'S UNIVERSITY

BMK Knowledge Resource Centre

Vithaldas Vidyavihar, Juhu Tara Road,
Santacruz (West) Mumbai - 400049

Optimisation-based weighted ensemble algorithm for predicting prices of spices (Record no. 131838)

MARC details
000 -LEADER
fixed length control field 02137nam a22001217a 4500
008 - FIXED-LENGTH DATA ELEMENTS--GENERAL INFORMATION
fixed length control field 250624b |||||||| |||| 00| 0 eng d
100 ## - MAIN ENTRY--PERSONAL NAME
Personal name Ankit Kumar Singh
245 ## - TITLE STATEMENT
Title Optimisation-based weighted ensemble algorithm for predicting prices of spices
300 ## - PHYSICAL DESCRIPTION
Extent P 776-784
520 ## - SUMMARY, ETC.
Summary, etc. biblio.abstract India plays a significant role in global agriculture production, including spices. Spices play an important role<br/>in cultural significance and economic trade relations<br/>while providing nutritional and medicinal benefits.<br/>The volatility and complexity of the price of spices require improved forecasting methods to support informed decision-making in agricultural markets. Recently,<br/>researchers have focused on using the traditional time<br/>series model as well as machine learning (ML) model<br/>to forecast the price of agricultural commodities. Using<br/>a standalone model struggles to capture the complex<br/>pattern in time series data. To overcome this challenge,<br/>ensemble machine learning approaches based on fixed<br/>weight (FW-ensemble) have been proposed. The ML<br/>models like artificial neural networks (ANN), random<br/>forest (RF), k-nearest neighbours (kNN), extreme gradient boosting (XGBoost), support vector regression<br/>(SVR) and the stochastic model, e.g. autoregressive integrated moving average (ARIMA) model have been<br/>used. The outputs of these models are ensembled using<br/>optimised fixed weights. In this study, the prices of two<br/>important spices, namely turmeric and coriander, from<br/>2010 to 2024, collected from AGMARKNET (https://<br/>agmarknet.gov.in/) were considered. The MCS algorithm was used to select the better-performing model.<br/>The empirical performance of the ensemble method<br/>was compared with that of the stochastic model<br/>(ARIMA), ML techniques (ANN, RF, kNN, XGBoost,<br/>SVR) and deep learning techniques, e.g. long shortterm memory (LSTM) and gated recurrent unit<br/>(GRU), based on several accuracy measures. It revealed that the FW-ensemble approach significantly outperformed the other candidate models in terms of<br/>prediction accuracy.
654 ## - SUBJECT ADDED ENTRY--FACETED TOPICAL TERMS
Subject <a href="Accuracy">Accuracy</a>
-- <a href="ensemble">ensemble</a>
-- <a href="forecasting">forecasting</a>
-- <a href=" spices"> spices</a>
-- <a href="machine learning">machine learning</a>
942 ## - ADDED ENTRY ELEMENTS (KOHA)
Koha item type Journal Article
Holdings
Withdrawn status Lost status Source of classification or shelving scheme Damaged status Not for loan Location (home branch) Sublocation or collection (holding branch) Date acquired Koha issues (times borrowed) Piece designation (barcode) Koha date last seen Price effective from Koha item type
    Dewey Decimal Classification     SNDT Juhu SNDT Juhu 24/06/2025   JP673.2 24/06/2025 24/06/2025 Journal Article