User:Ptolemez/MyBidMax

myBidMax™
myBidMax™ is a software product created, developed and designed by Reputronix a Silicon Valley base company, as a tool to assist eBay sellers and eBay buyers before and during their auctions.

myBidMax™ uses state-of-the-art Computing Technology to collect (data mining) and interpret large scale amounts of online information into very simple, user friendly terms. eBay users when using this tool are then in a position to be much more knowledgeable and productive in their online eBay business.

myBidMax™ is a predictive machine capable of predicting the various states of eBay auctions seen from the seller or buyer side.

An eBay auction state is defined by the following parameters:
 * Start Bid
 * Reserve Bid
 * Current Bid
 * Time left until the end of auction
 * Number of bidders
 * Number of bids

to which is added Target Bid a user provided input and Expected bid a computed value.

System Architecture
myBidMax™ consists of the following subsystems:
 * 1) Machine Learning Engine - This is an in-house component and is not part of the Software distribution but rather its output data as generated predictive models is part of the distribution and referred to as the definition of the system.
 * 2) Knowledge Base contains the set of rules necessary to process user input along with the Inference Engine
 * 3) Inference Engine - This is the component used to infer from auction data the predicted states of a particular auction
 * 4) File Manager - this component is responsible for maintaining user provided input such as auction lists and the data acquired from eBay auction item
 * 5) User Interface responsible for providing the mean for creating auction lists and presenting the results displayed in a comprehensible language

Background and History
myBidMax™ is the brainchild of Ninon Crane and Jean Claude Lemoyne. In 2005 after being for a while a seller on eBay Ninon Crane made the remark to Jean Claude Lemoyne a Software Engineer that there was not real guide to pricing auction items on eBay. It was then decided to create a software product that will do exactly that. Initially designed for sellers it became quite obvious that the same approach will help buyers. And the project was born. It will take several months of design to come up with a clear solution. The concept was to try to generate predictive models based on Machine Learning techniques. Training of the system will be accomplished by providing it with eBay data collected from the internet. The first phase saw the assembling of large scale amounts of data by data mining the World Wide Web. After transforming the collected data into database records suitable for programing. These programs then generated mathematical models which then were tested for their performance. It's only when the results where more than satisfactory and later on after tuning up the models, the results obtained had a percent accuracy between 88% to 98% that it was decided to go ahead and build myBidMax™. It will take to years to finally complete the system. It was launched on May 27, 2009.

eBay Auctions Predictive Tool
Jean-Claude Lemoyne an Operations Research specialist recalled a conversation he had with a plant manager about optimizing plant production. This manager when asked what he would need to optimize his production he replied: "give me an accurate demand forecast and I will do the rest". In dead indivual Auctions may be modelled as supply-demand models; the sellers representing the supply-side and the buyers the demand-side except that auctions are too discrete of a model to be entirely described in terms of supply-demand. Taken as a whole certainly eBay auctions and auctions in general will not escape supply-demand laws. Nonetheless Auctions are microcosm economies where supply-demand applies. First of all myBidMax™ is a tool that predicts demand. It provides myBidMax™ users with the most accurate price projection given some basic parameters. Naturally these projections are based on the models generated by the system. This information referred to as the system definition needs to be refreshed periodically. So myBidMax™ is a decision tool also designed to run if-scenarios. The user may vary the number of bids or may set a target bid to reach. myBidMax™ reacts by recalculating the odds of such variations. This tool takes into account the dynamic nature of the bids. As auctions are being conducted and before being completed users have the ability to refresh by the click of a button their auction states. The system will recalculate the odds for reaching a target bid price or simply by re calculating the new expected bid.

Terms Used

 * 1) Auction
 * 2) eBay
 * 3) Auction Theory
 * 4) Auction Algorithm
 * 5) Auction Betting Exchange
 * 6) Operations Research
 * 7) Predictive
 * 8) Computing Technology
 * 9) Predictive Modelling
 * 10) Predictive Classification
 * 11) Machine Learning
 * 12) Knowledge Base
 * 13) Probability Theory
 * 14) Mathematical Models