What question did this study set out to answer?

This research aims to improve the accuracy of monthly sales forecasting using an advanced machine learning model.

May 31, 2026Open Access

Seasonality and Trend in Sales Forecasting

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial develops a novel sales forecasting model in retail, highlighting its effectiveness and adaptability.

This research aims to improve the accuracy of monthly sales forecasting using an advanced machine learning model.
Developed a forecasting model integrating time series decomposition and gradient boosting techniques.
Utilized Pearson and Spearman correlation tests to analyze data structure and engineered features based on sales history and local indicators.
Evaluated model performance using real operational sales data from a Kazakhstani retail business.
LightGBM achieved an R² of 0.81, explaining 81% of total sales variance with an RMSE of 1,996.
Feature importance analysis showed incoming stock volume, product category, and selling price outperformed temporal features in predictive power.
The model is designed to incorporate local events and is adaptable for various markets.

Bibinur et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

May 31, 2026Open Access

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial develops a novel sales forecasting model in retail, highlighting its effectiveness and adaptability.

This research aims to improve the accuracy of monthly sales forecasting using an advanced machine learning model.
Developed a forecasting model integrating time series decomposition and gradient boosting techniques.
Utilized Pearson and Spearman correlation tests to analyze data structure and engineered features based on sales history and local indicators.
Evaluated model performance using real operational sales data from a Kazakhstani retail business.
LightGBM achieved an R² of 0.81, explaining 81% of total sales variance with an RMSE of 1,996.
Feature importance analysis showed incoming stock volume, product category, and selling price outperformed temporal features in predictive power.
The model is designed to incorporate local events and is adaptable for various markets.

Bibinur et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark