Wavelet decomposition in signal processing has been widely used in the literature. The popularity of machine learning (ML) algorithms is increasing day by day in agriculture, from irrigation scheduling and yield prediction to price prediction. It is quite interesting to study wavelet-based stochastic and ML models to appropriately choose the most suitable wavelet filters to predict agricultural commodity prices. In the present study, some popular wavelet filters, such as Haar, Daubechies (D4), Coiflet (C6), best localized (BL14), and least asymmetric (LA8), were considered. Daily wholesale price data of onions from three major Indian markets, namely Bengaluru, Delhi, and Lasalgaon, were used to illustrate the potential of different wavelet filters. The performance of wavelet-based models was compared with that of benchmark models. It was observed that, in general, the wavelet-based combination models outperformed other models. Moreover, wavelet decomposition with the Haar filter followed by application of the random forest (RF) model gave better prediction accuracy than other combinations as well as other individual models.
15 February 2023 in Intelligent Systems with Applications
Accurate prediction of time series data is crucial for informed decision-making and economic development. However, predicting noisy time series data is a challenging task due to their irregularity and complex trends. In the past, several attempts have been made to model complex time series data using both stochastic and machine learning techniques. This study proposed a CEEMDAN-based hybrid machine learning algorithm combined with stochastic models to capture the volatility of weekly potato price in major markets of India. The smooth decomposed component is predicted using stochastic models, while the coarser components, selected using MARS, are fitted into two different machine learning algorithms. The final predictions for the original series are obtained using optimization techniques such as PSO. The performance of the proposed algorithm is measured using various metrics, and it is found that the optimization-based combination of models outperforms the individual counterparts. Overall, this study presents a promising approach to predict price series using a hybrid model combining stochastic and machine learning techniques, with feature selection and optimization techniques for improved performance.
18 March 2023 in International Journal of Environment and Climate Change
A novel method for rainfall forecasting has been proposed using Multi Resolution Analysis (MRA). This approach decomposes annual rainfall series and long-term climate indices into component sub-series at different temporal scales, allowing for a more detailed analysis of the factors influencing annual rainfall. Multiple Linear Regression (MLR) is then used to predict annual rainfall, with climate indices sub-series as predictive variables, using a step-wise linear regression algorithm. The proposed model has been tested on Indian annual rainfall data and compared with the traditional MLR model. Results show that the MRA-based model outperforms the traditional model in terms of relative absolute error and correlation coefficient metrics. The proposed method offers several advantages over traditional methods as it can identify underlying factors affecting annual rainfall at different temporal scales, providing more accurate and reliable rainfall forecasts for better water resource management and agricultural planning. In conclusion, the MRA-based approach is a promising tool for improving the accuracy of annual rainfall predictions, and its implementation can lead to better water resource management and agricultural planning.
Expensive black-box optimization problems, where each function evaluation requires costly simulations or experiments, demand algorithms that balance solution quality with computational efficiency. This paper introduces the Canine Olfactory Optimization (COO) algorithm, a surrogate-assisted metaheuristic inspired by canine olfactory behavior. Dogs possess approximately 300 million olfactory receptors and devote significant neural capacity to scent processing, enabling gradient tracking, spatial scent mapping, and coordinated pack hunting: mechanisms that motivate the proposed optimization framework. COO integrates established techniques within a unified architecture, combining adaptive multi-population search, ensemble-based surrogate modeling, gradient-based local refinement, and a hysteresis-controlled surrogate activation strategy. Rather than proposing entirely new operators, COO emphasizes structured coordination to stabilize surrogate usage and balance exploration-exploitation dynamics. Two complementary studies evaluate performance. Study 1 compares COO with Differential Evolution, Particle Swarm Optimization, Grey Wolf Optimizer, and Whale Optimization Algorithm across 24 benchmark variants (6 functions × 4 transformations: base, shifted, rotated, shifted-rotated) using 10 independent seeds. Study 2 examines artificial neural network hyperparameter optimization on four regression datasets. COO achieves the best efficiency score on the shifted-rotated Schwefel function, a deceptive landscape with distant global optima. It attains the highest observed transformation robustness scores (1.11 for solution quality and 1.42 for efficiency), with modest differences among algorithms. On the Diabetes dataset, COO reduces test MSE by 2.28% relative to DE (2918 vs. 2986). Pareto analysis indicates favorable accuracy-efficiency trade-offs. While not universally dominant, COO demonstrates particular suitability for expensive multimodal black-box problems.
18 February 2026 in Intelligent Systems with Applications