Finding the largest consensus set is one of the key ideas used by the original RANSAC for removing outliers in robust-estimation. However, because of its random and non-determinis...
Let K be a polytope in Rn defined by m linear inequalities. We give a new Markov Chain algorithm to draw a nearly uniform sample from K. The underlying Markov Chain is the first t...
— Small gates, such as AND2, XOR2 and MUX2, have been mixed with lookup tables (LUTs) inside the programmable logic block (PLB) to reduce area and power and increase performance ...
This a summary of the author's PhD thesis supervised by Leo Liberti, Philippe Baptiste and Daniel Krob and defended on 18 June 2009 at Ecole Polytechnique, Palaiseau, France. ...
We propose a novel boosting algorithm which improves on current algorithms for weighted voting classification by striking a better balance between classification accuracy and the ...