Forecasting Market’s Demand And Supply With Machine Learning And Local Weather

The study explores the use of Artificial Intelligent Machine Learning (AI/ML) models in forecasting Small and Medium Scale Enterprises (SMEs’) local goods’ quantity purchase, buying and goods selling prices, from the day-to-day local weather conditions of low and high temperatures. The use of weather parameters in forecasting for SME market parameters proved to be an accurate procedure for SME market parameters’ forecasting than other methods which use time-series methods. The weather was chosen as the independent variable, as it impacts local production and influences human behaviour in ways that most traditional methods cannot compute. With the inclusion of weather, AI/ML forecast models considered the seasonal variations of demand and supply traits due to seasonal shifts in the popularity of certain goods and services. This can give SMEs a better market business insistence. SMEs herein included the marketers who trade in household edible commodities. The study’s findings will give SMEs an added advantage, with insists on increasing their savings, reducing losses, minimizing resource wastage, with a prior overview of coming business risks and opportunities. The AI/ML forecasting models explored, included the Multivariate Linear Regression model, and Logistic Regression model built using visual C# DotNet, and an ‘Encoder-Decoder Long Short-Term Memory (LSTM) Neural Network modelled with python language. The study contributes to the body of knowledge by elaborating how ML applications will give SMEs a chance to take pre-emptive measures, by prior planning and controlling cash flows, balancing spending with revenues and adjusting sales targets. It also proposes how ML applications should use personalized localized data in forecasting business parameters

REFERENCES

[1] Bevans, R. (2020, July 16). The p-value explained. Scribbr. Retrieved from www.scribbr.com/statistics/p-value
[2] Bousquet, O., Boucheron, & Lugosi, G. (2003). Introduction to Statistical Learning Theory. Advanced Lectures on Machine Learning. ML 2003. Lecture Notes in Computer Science, vol 3176. Springer, Berlin, Heidelberg. Retrieved from https://doi.org/10.1007/978-3-540-28650-9_8
[3] Carabello, F. (2019, June 4). Market Futures: Introduction to Weather Derivatives. Investopedia. Retrieved from https://www.investopedia.com/trading/market-futures-introduction-to-weather-derivatives/
[4] Chibwe, F. (2014). THE RELATIONSHIP BETWEEN INFLATION AND ECONOMIC GROWTH IN ZAMBIA (1980-2011):. LUSAKA: University of Zambia.
[5] Duncan, R., & García, E. M. (2018, June 8). As good as a random walk: Inflation forecasting in emerging market economies. CEPR Policy Portal. Retrieved from https://voxeu.org/article/inflation-forecasting-emerging-market-economies
[6] Editor. (2012, November 5). The meteorological department challenged to improve their communication. LusakaTimes.com. Lusaka.
[7] Fritsch, D. (2015, October 2). 5 Demand Planning Challenges Facing Distributors Today. Retrieved from https://www.eazystock.com/blog/5-demand-planning-challenges-facing-distributors-today/?cn-reloaded=1
[8] IBM Corporation. (2017). Bayesian Related Sample Inference: Normal (25.0.0 ed.). Retrieved from https://www.ibm.com/docs/en/spss-statistics/25.0.0?topic=statistics-bayesian
[9] Kelber, J. (2020, February 06). The Top Challenges in Supply Chain Forecasting. Retrieved from https://blog.flexis.com/top-challenges-supply-chain-forecasting
[10] Mester, L. (2016, December 05). Recent Inflation Developments and Challenges for Research and Monetary Policymaking. The 47th Konstanz Seminar on Monetary Theory and Monetary Policy. Insel Reichenau, Germany. Retrieved from https://www.clevelandfed.org/en/newsroom-and-events/speeches/sp-20160512-recent-inflation-developments-and-challenges-for-research-and-monetary-policymaking.aspx