Don Reba

Spain

Age: 30

Representative of the Holy Order in the Arkanarian Province. Bishop and Councilor of War, servant of Our Lord!

http://careers.stackoverflow.com/alexey-badalov

1d
comment Advantages of SIFT over HOG
It depends on what you need the features for. Like you said, they are different kinds of features used for different purposes.
1d
comment Evaulating Algorithm and decision making system
put on hold as too evil
Dec
10
comment What is the best Microsoft's language for Artificial Intelligence projects?
It's the most appropriate language for all things.
Dec
10
comment What is the best Microsoft's language for Artificial Intelligence projects?
C++, of course.
Dec
9
awarded Enlightened
Dec
9
awarded Nice Answer
Dec
9
answered Where am I in my array?
Dec
7
comment Efficient algorithm to determine running time
This is not actually a question.
Dec
7
comment Using Vectors in C++?
vec.push_back(ch.data()) except you won't be able to print it to cout, because your char array is not null-terminated.
Dec
6
answered Q learning computation: states unknown
Dec
5
comment How do you implement quicksort with the pivot in the middle position and what is this variation called?
@GeneralDynamix, the Wikipedia article on Quicksort linked in this answer addresses this.
Dec
5
answered Deadlock when using condition variables for two way communication
Dec
4
comment Is Q-Learning Algorithm's implementation recursive?
I call t+1 the current state, because this is the one that was reached upon receiving the reward.
Dec
4
answered Is Q-Learning Algorithm's implementation recursive?
Nov
23
comment Machine learning : learn feature value range for a classification
Isn't cheapness entirely determined by price? If you want to estimate appropriate price for a product from its features — that is a different problem.
Nov
23
answered Can we find the redundant features in original data after applying PCA
Nov
23
revised This program for swapping two halves of an array is not working
un-double-space
Nov
23
comment Machine learning : learn feature value range for a classification
Maybe I'm missing something, but why not simply find under which price lie 90% of the cheap products in your sample? It sounds like a trivial data exploration problem, not a case for machine learning.
Nov
19
awarded Popular Question
Nov
19
comment Reinforcement learning for continuous states, discrete actions. Algorithms?
An optimal policy would map every observation to the action that yields maximum expected reward in the long run. Meaning, specific states are mapped to specific actions, not distributions of actions, where some are taken merely more often than others.
1 2 3 4 5