Data Mining - Stochastic Gradient Boosting Tree(repost)

Realization of Stochastic Gradient Boosting Tree algorithm for arbitrary number of parameters. ## Deliverables <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="134">The essence of <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="135">the project</span></span> is the realization of Stochastic Gradient Boosting Tree algorithm for arbitrary number of parameters. Expecting data is <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="645">data</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="646">with a large</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="647">number</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="648">of observations</span> (up to 100 000) <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="649">and</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="650">a small</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="651">number of</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="652">variables (up to 100)</span><span title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="653">. </span> <span class="hps" id="result_box" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="577"><span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="428">Source data f</span><span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="429">or the</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="430">creating of</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="431">the tree, created tree, data to</span></span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="580">analyze and</span> <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="583">prediction results </span>should be stored in database. Planning OS - UNIX (FreeBSD preferrable). Languge - C++, SQL. I'm waiting for: 1) E<span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="773">stimate <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="774">the required</span> server <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="775">capacity (max acceptable for me is the Quad Core CPU & 4GB RAM). </span></span>2) Development of the appropriate database structure. 3) <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="618">Choosing the right <span class="hps" title="???????, ????? ??????? ?????????????? ???????" closure_uid_2l4fds="619">database</span></span> (I think it should be PostGre SQL, but it is discussable). 4) Realization of the algorithm. 5) Testing for correct work and estimated performance. Please do not hesitate to contact me if you have any questions.