Tesseract

Tesseract is a library for performing diverse feature selection for linear regression problems using spectral regularizers.

Instruction

0. install the dependencies (mentioned below)
1. create a directory named tests in the root directory
2. run "make" for compiling lib, unit/integration tests (goes inside tests dir) and documentation
3. run "make check" for running the tests

Dataset

• MNIST dataset is obtained from here
• Boston Housing dataset is obtained from here

Dependencies

• Eigen3 linear algebra library
• GCC C++ compiler with C++11 features enabled (4.8.3 or higher)
• Doxygen