Information Technology Reference
In-Depth Information
2.2
Probability Calculating
The Bayesian network classifier is a simple classification method, which classifies a
case by determining the probability of it belonging to the i-th target category Yi. As
investors, we main concern is the loans that will pay back with high probability. These
probabilities are calculated as
(
)
j
j
j
n
P
BStatus
=
1
|
X
=
x
,
X
=
x
,
,
X
=
x
1
1
2
2
n
(
)
j
j
j
n
P
BStatus
=
1
X
=
x
,
X
=
x
,
,
X
=
x
1
1
2
2
n
=
(
)
j
j
j
n
P
X
=
x
,
X
=
x
,
,
X
=
x
1
2
n
1
2
(
)
n
(
)
j
k
=
P
BStatus
=
1
P
X
=
x
|
parrent
(
X
)
k
k
k
1
(2)
=
(
)
j
j
j
n
P
X
=
x
,
X
=
x
,
,
X
=
x
1
2
n
1
2
3
Experiment Results and Comparison
3.1
Dataset
The dataset used in our experiments is from Prosper.com. Prosper includes six rela-
tional data tables, which are Members, Groups, Credit Profile, Listings, Loans and
Bids data tables. The Listing table is the most important for our modeling. A Listing is
created by a Borrower to solicit bids by describing themselves and the reason they are
looking to borrow money. If the Listing receives enough bids by Lenders to reach the
Amount Requested then after the Listing period ends it will become a Loan.
In our experiments we use seven attributes from the Listing table, which are de-
scribed in details below.
AmountRequested The amount that the member requested to borrow.
BorrowerRate The rate is computed as the LenderRate + GroupLeaderRewardRate
(if applicable) + BankDraftFeeAnnualRate (if applicable).
CreditScore The credit score of the borrower at the time the listing was created
Category The Category is one of the following numerical values : 0 Not available, 1
Debt consolidation, 2 Home improvement, 3 Business, 4 Personal loan, 5 Student use, 6
Auto, 7 Other.
DebtToIncomeRatio The debt to income ratio of the borrower at the time the listing
was created.
IsBorrowerHomeowner This attributes specifies whether or not the member is a
verified Homeowner.
BidCount The total number of Bids.
3.2
Data Preprocessing
Bayesian nodes deal with discrete data, however, only category (0~7) and IsBorro-
werHomeowner (0=false, 1=true) are discrete, the others are continuous values.
 
Search WWH ::




Custom Search