Authors:
R. Princy Reshma, S. Deepak, S. R. M. Tejeshwar, P. Deepika, Muhammad Saleem

Addresses:
1,2,3,4Department of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. 5Department of Management and Economics, Kunming University of Science and Technology, Kunming, Yunnan, China. pr0695@srmist.edu.in1, ds1696@srmist.edu.in2, gg4307@srmist.edu.in3, deepikap2@srmist.edu.in4, m.saleem647@gmail.com5

Abstract:

Fast-paced internet auctions require smart choices and real-time bidding. Both buyers and sellers in online auctions need price prediction. Accurate estimates let suppliers list products at the right time and type, and bidders choose. Data-driven decision-making requires advanced analytics. Selling prices are calculated first because auctions are dynamic markets with uncertain bid-by-bid evolution. Traditional papers focus on final sale prices, although interim bidding can aid. Strategic knowledge, value perception, and market enthusiasm drive bids. Bidder values and market interest determine bidding. The entire auction price prediction system that shows sophisticated bid history in this research fills this gap. Use a certified online auction dataset. To predict the auction form's eventual price and buyer bids, imported data must be cleansed and preprocessed to remove noise. Strong Z-scores eliminate outliers, ensuring bid progression analysis accuracy. Auction data comprises item, length, and open price. K-Fold cross-validation improves models across datasets. Assessors use bid forecast precision measure RMSE. The statistical basis predicts accurately, and sophisticated visualizations help stakeholders understand auction behavior. Finally, the bids are compared to historical data from auctions with the same number of bids to validate the system. To assure auction behavior, the paper evaluates predictions and bids. The bid value standard deviation is used to evaluate bids. These phases compare bids to auction history. Bids are visualized against auction data. Box and line graphs are popular. Compare produced bids to past auction data to understand bid distribution and dynamics. Auction players gain from its real-time predictions over other systems. The paper's holistic approach beats auction analytics. This paradigm provides a complete toolkit for forecasting bid-by-bid dynamics, interpretation, and visualization, not selling prices. 

Keywords: Root Mean Squared Error (RMSE); Online Auctions; Predictive Analytics; XG Boost; Linear Regression; Bid Progression; K-Fold Validation; Data-driven Decision-Making; Machine Learning; Strategic Bidding; Bid Trajectory Forecasting.

Received on: 27/12/2022, Revised on: 19/02/2023, Accepted on: 03/06/2023, Published on: 15/12/2023

FMDB Transactions on Sustainable Technoprise Letters, 2023 Vol. 1 No. 2, Pages: 106-122

• Views : 189
• Downloads : 5

Download PDF