Analysis of Real-Time Data Using AI: Future Sales Prediction

Analysis of Real-Time Data Using AI: Future Sales Prediction

Sivasankari Jothiraj, P. Divya Bharathi, B. R. D. Rigveda, K. Aksharaa, S. Sabreen Safira
Copyright: © 2024 |Pages: 29
DOI: 10.4018/979-8-3693-4276-3.ch006
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

In the realm of advertising, predicting future sales is a paramount concern for businesses seeking to optimize their marketing budgets. This chapter outlines a research study that employs a linear regression model to forecast sales trends for three traditional advertising channels: TV, newspaper, and radio. The study begins by gathering historical data on sales, advertisement spending, and other relevant variables for these advertising channels. Utilizing this data, a linear regression model is constructed to recognize the connections between advertising expenditures and sales performance. By examining the historical performance of these channels, the research seeks to uncover insights into how advertising budgets influence sales outcomes. The research aims to provide advertisers, marketers, and businesses with a predictive tool for optimizing their advertising strategies and budgets. Ultimately, this study equips advertisers and stakeholders with a quantitative framework to enhance their strategic planning.
Chapter Preview
Top

Literature Survey

Development of a Fourier Series Forecasting Model for Predicting the Sales Volume Selected Manufacturing Company

The main goal pertaining to this research project is to establish a forecasting model capable of predicting future product sales that experience seasonal fluctuations. These variations in sales volume are attributed to changes in consumer demand for the product. The objective of this research is to create a mathematical forecasting model to forecast future sales projections volume of a plant or production company. The framework is derived by combining a Fourier series of cosine functions with a linear equation. This integrated approach is valuable for forecasting the future sales volume of products or goods characterized by sinusoidal fluctuations in sales. The model is tailored for products characterized by seasonal in demand (D.H. Oladebeye & O.S. Ejiko, 2015).

Intelligent Sales Prediction Using Machine Learning Techniques

This paper highlights the comparative analysis of the prediction performance of three algorithms. According to the performance metrics, the Gradient Boost Algorithm demonstrates the highest overall accuracy at 98%. The second-ranking algorithm is the Decision Tree Algorithms, which achieves nearly 71% overall accuracy, followed by the Generalized Linear Model with a precision of 64%. Owing to the considerable execution time and the intricate nature of handling an extensive set of records, certain records were omitted during the analysis phase. Concurrently, the fields and attributes utilized in this analysis were found to be insufficient for conducting further analysis (Sunitha Cheriyan, Shaniba Ibrahim et al., 2018).

Complete Chapter List

Search this Book:
Reset