Andreas

191 reputation

bio	website
	location
	age
visits	member for	1 year, 5 months
	seen	Apr 10 at 7:24
stats	profile views	15

	191 reputation	bio	website		visits	member for	1 year, 5 months
7 badges		location			seen	Apr 10 at 7:24

summary answers questions tags badges favorites bounties reputation activity

41 Actions

suggestions reviews revisions comments badges posts accepts all

Jan 13	asked	Problem with singular covariance matrices when doing gaussian process regression
Jan 12	comment	Gaussian Process covariance matrix gets zero determinant The inputs in a gaussian process are some training points and normally we set the X* points over the inputs. Thus we have more X* points than X points and they span the same range. Maybe this couls lead sometimes to the decribed X=AX* + B behavior
Jan 12	comment	Gaussian Process covariance matrix gets zero determinant The inputs in a gaussian process are some training points and normally we set the X* points over the inputs. Thus we have more X* points than X points and they span the same range. Maybe this couls lead sometimes to the decribed X=X* behavior
Jan 11	comment	Gaussian Process covariance matrix gets zero determinant I'm using the RBF covariance function. Squared exponention. When i exclude the x* covariances from the matrix everything is fine as well. It just happens in some cases. I guess it is really a problem with the implementation of the Commons Math solver.
Jan 11	comment	Gaussian Process covariance matrix gets zero determinant It seems to be a problem with the Commons.Math Matrix LU Solver. But i guess those Implementations only react strange at some sizes of the covariance matrix. If i add one more training point everything is fine again.
Jan 11	asked	Gaussian Process covariance matrix gets zero determinant
Jan 10	comment	Gaussian process scale targets Thank you for the answer. I forgot to optimize sigma_n. I only optimized sigma_f. This caused the strange behavior. But your displayed thoughts are very useful. Thanks
Jan 10	awarded	Scholar
Jan 10	accepted	Implementation of Gaussian process
Jan 10	accepted	Gaussian process scale targets
Jan 10	revised	Gaussian process scale targets edited body
Jan 10	asked	Gaussian process scale targets
Jan 4	comment	Implementation of Gaussian process Additionally my predicted curve isn't getting any smoother as i adjust lenght scale. The prediction of the target points always stays the same whereas the prediction of the unknown points gets smoother. But i want the "interpolation" of the given points to be smooth as well when i set a large length scale
Jan 4	awarded	Supporter
Jan 4	awarded	Editor
Jan 4	revised	Implementation of Gaussian process added code for implementation
Jan 4	comment	Implementation of Gaussian process I just experimented with differnent length scales. The result is that i can influence how fast i decline to the zero values. What i expect though is a behavior like the java applet by Andreas Geiger rainsoft.de/projects/gausspro.html. When there is a target which is in the near of a local maximum the maximum is drawn by the process before its declining. I'm using the RBF function as well but in my implementation the prediction is always declining. Mybe my implementation of the left matrix devision is faulty. But i just invert and multiply which should give the right result.
Jan 4	asked	Implementation of Gaussian process
Dec 12	comment	Assessing quality of similarity measure No i selected differnt similarities that fit the matching idea logically. My problem is to sub-select attributes from the data that fully determine their class (Its a word distribution where only a few words have high probability and the rest i a low probability long tail). Hence i subselect differnt amounts of those words and compare the outcomes of the similarity measure. The target would be an optimal seperation between matches and non matches. In the moment i only look at how the values spread and the optimisation target is to max this spread
Dec 12	awarded	Student