Prev: How to have drop down menu to select in inputdlg
Next: Cos infinite series
From: Yerkebulan Shalmaganbetov on 21 Mar 2010 19:52
Pattern discovery in financial data provided in company balance sheets
for evaluation and prediction of credit risk and survivability.

The project wil involve te analysis and processing of financial
records and balance sheets of medium to large enterprises, with the
aim to:
a) model company's financial state and extract statistical rules that
govern incurring characteristic patterns, using learning pattern
recognition schemes, eg. neural-based methods or Bayesian statistics.
This will result in a generic model for describing credit risk in bank
financing.

b) develop an algorithm to effectively detect patterns of high risk
and low survivability.

Implementation will be done preferably in the Matlab environment.
From: Walter Roberson on 21 Mar 2010 20:34
Yerkebulan Shalmaganbetov wrote:
> Pattern discovery in financial data provided in company balance sheets
> for evaluation and prediction of credit risk and survivability.

> b) develop an algorithm to effectively detect patterns of high risk
> and low survivability.

That's a *lot* more than 3 days work!!

The group I work for has put more than 20 person-years into developing
robust data classification algorithms (for medical purposes, but the
programs wouldn't care if it was fed time series of monetary values.)
Unless you find someone who is *already* doing this work and *already*
has a code library, there is no realistic hope that they would be able
to finish such a project by the 24th.
 | 
Pages: 1
Prev: How to have drop down menu to select in inputdlg
Next: Cos infinite series