Page 281 - Data Science Algorithms in a Week

262 Oloruntomi Joledo, Edgar Gutierrez and Hatim Bukhari

and company website. The case study helps to select and define boundaries and core
areas of interactions on the platform.
To describe the online consumer-to-consumer (social) lending in the context of an
ecommerce system, liquidity, pricing models, and uncertainty, hybrid modeling is used.
Growth in the industry partly results from investors being discouraged by stock market
returns and lower interests provided by banks. Results from business case studies and
literature review indicate that the success of peer-to-peer (P2P) lending business process
innovation has not been proven. As an example, it is beneficial to balance the number of
lenders and qualified borrowers that can effectively meet the mutual needs of the
customers. Because this form of lending is insecure, lenders are exposed to a risk of
default by the borrower. The platforms have to deal with uncertainties that pervade all
aspects of its operations. Any unplanned downtime, outage or system hack can have long
term effects to its operations and credibility (Joledo et al. 2014).

MAPPING OF THE FRAMEWORK TO A HYBRID SIMULATION SYSTEM

The hybrid simulation models in this research are developed using AnyLogic
(http://www.anylogic.com/). AnyLogic has the capability of creating mixed discrete-
continuous simulations of ABS and SD models in the same interface. Seller consumers
come into the system with goods to sell. These consumers require different thresholds of
returns. Buyers have different cut-off prices which they can pay for transactions on the
platform. Consumers are modeled as agents whose behaviors elicit corresponding
responses. The dynamics of price agreement are also modeled in the agent-based system.
The environment is modeled in SD with agents living therein. The population of
consumers is disaggregated to individual level using agents.
In the simulation of business processes, interactions between players are modeled
using statecharts. The system is simulated over a period of eight years to gain insights in
to the behavior of participants and how their individual or collective actions affect the net
income and in turn the profit margin. The output of the overall system is viability
measured by the default rates, net income and profit margin. Outputs of the ABS
subsystem are fed into the time-continuous SD model (strategic layer). The assumption
for this study is that the seller seeks to sell his product at a profit while the buyer seeks to
pay the minimum cost for a particular product. The provider supplies a medium for the
realization of customer utility while making a profit in the process.