| 1 | /*! |
|---|
| 2 | \mainpage Bayesian Decision-Making toolbox for C++ |
|---|
| 3 | |
|---|
| 4 | \version 0.1 |
|---|
| 5 | |
|---|
| 6 | \author Vaclav Smidl |
|---|
| 7 | |
|---|
| 8 | BDM is a collection of methods for selected tasks of Bayesian decision-making, such as estimation, filtering and control. |
|---|
| 9 | |
|---|
| 10 | \section fea Features |
|---|
| 11 | At present the following algorithms are implemented: |
|---|
| 12 | - \b Bayesian \b filtering : Kalman filter, EKF, patricle filter (PF), |
|---|
| 13 | - these can be combined mutualy together in mode demanding schemes, see marginalized particle filter MPF |
|---|
| 14 | |
|---|
| 15 | - \b Classification using mixtures of exponential famiuly models (MixEF), |
|---|
| 16 | - \b Density \b estimation : using mixtures (MixEF), density composition (merger) |
|---|
| 17 | |
|---|
| 18 | \section down Download and Use |
|---|
| 19 | The library is available under GPL, see installation instructions on page \ref install |
|---|
| 20 | |
|---|
| 21 | Precompiled Mex files for use within Matlab are available \ref mexfiles |
|---|
| 22 | |
|---|
| 23 | \section app Design Approach |
|---|
| 24 | The toolbox is designed using \b object-oriented approach with two design criteria: |
|---|
| 25 | \li relation to mathematics, |
|---|
| 26 | \li efficient evaluation, |
|---|
| 27 | while the first one is more important than the latter. |
|---|
| 28 | |
|---|
| 29 | Hence, each mathematical object such as probability density is |
|---|
| 30 | represented by one software object. The resulting algorithms are |
|---|
| 31 | then implemented as operations on these objects. In cases when |
|---|
| 32 | more efficient solution can be achived when this structure is not respected, |
|---|
| 33 | a parallel implementation is created and clearly marked as specific. |
|---|
| 34 | |
|---|
| 35 | OpenMP is used to achive efficient implementation on parallel architectures. |
|---|
| 36 | |
|---|
| 37 | \section impl Implementation |
|---|
| 38 | |
|---|
| 39 | BDM is build on top of \c IT++ which wraps numerically efficient operations of linear algebra into easy to use C++ classes. Thanks to this excellent library, writing of numerical algorithms is as easy as in Matlab but we gain significant advantages: |
|---|
| 40 | |
|---|
| 41 | \li computational speed comparable to built-in Matlab function, and surpassing interpreted Matlab in order of magnitudes, |
|---|
| 42 | \li native support for object-oriented programming, |
|---|
| 43 | \li support for templates which is often more appropriate than object-oriented programming, |
|---|
| 44 | \li cross-platform compatibility. |
|---|
| 45 | |
|---|
| 46 | |
|---|
| 47 | */ |
|---|
