Template:StimulusTaskParams - Revision history

Mellinger: /* MinimumEvidence */

2023-11-23T15:46:01Z

MinimumEvidence

← Older revision		Revision as of 15:46, 23 November 2023
Line 58:		Line 58:
	The meaning of the actual number entered into the ''MinimumEvidence'' parameter is relative to the amount of within-class variance present in the classification score. An evidence of 0 means a 50:50 chance for correct classification. Increasing the evidence value by two standard deviations corresponds to an improvement by a factor of roughly 88:12, four standard deviations correspond to (88:12)^2=(98:2) ... etc, approaching perfect classification as evidence increases towards infinity.		The meaning of the actual number entered into the ''MinimumEvidence'' parameter is relative to the amount of within-class variance present in the classification score. An evidence of 0 means a 50:50 chance for correct classification. Increasing the evidence value by two standard deviations corresponds to an improvement by a factor of roughly 88:12, four standard deviations correspond to (88:12)^2=(98:2) ... etc, approaching perfect classification as evidence increases towards infinity.

	In classifier training, classifier weights may be normalized such that within-class variance is 1 (this is done by recent versions of the P300Classifier tool). In this case, you may use the following equations to convert between the ''MinimumEvidence'' parameter <math>\eta</math>, and the correct classification chance <math>\chi</math>:		In classifier training, classifier weights may be normalized such that within-class variance is 1 (this is done by BCI2000Analysis, and recent versions of the P300Classifier tool). In this case, you may use the following equations to convert between the ''MinimumEvidence'' parameter <math>\eta</math>, and the correct classification chance <math>\chi</math>:

	<math>\chi = \frac{50}{50} \left(\frac{88}{12}\right)^{\eta/2}</math>.		<math>\chi = \frac{50}{50} \left(\frac{88}{12}\right)^{\eta/2}</math>.

Mellinger: /* PhotoDiodePatch */

2022-03-30T16:23:41Z

PhotoDiodePatch

← Older revision		Revision as of 16:23, 30 March 2022
Line 101:		Line 101:
	====PhotoDiodePatch====		====PhotoDiodePatch====
	Display a photo diode patch on the stimulus window.		Display a photo diode patch on the stimulus window.
	Recording from a photo diode located on that patch will allow triggering on actual stimulus delivery (see ~~parameter~~ OnsetExpression).		Recording from a photo diode located on that patch will allow triggering on actual stimulus delivery (see [[User_Reference:P3TemporalFilter#OnsetExpression]]).

	====PhotoDiodePatchHeight====		====PhotoDiodePatchHeight====

Mellinger at 16:16, 30 March 2022

2022-03-30T16:16:44Z

← Older revision		Revision as of 16:16, 30 March 2022
Line 98:		Line 98:

	'''NOTE:''' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [[Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|these notes on how to use a different classifier]].		'''NOTE:''' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [[Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|these notes on how to use a different classifier]].

			====PhotoDiodePatch====
			Display a photo diode patch on the stimulus window.
			Recording from a photo diode located on that patch will allow triggering on actual stimulus delivery (see parameter OnsetExpression).

			====PhotoDiodePatchHeight====
			Photo diode patch height in relative coordinates (between 0 and 1).

			====PhotoDiodePatchWidth====
			Photo diode patch width in relative coordinates (between 0 and 1).

			====PhotoDiodePatchLeft====
			Photo diode patch left position in relative coordinates (between 0 and 1).

			====PhotoDiodePatchTop====
			Photo diode patch top position in relative coordinates (between 0 and 1).

			====PhotoDiodePatchShape====
			Photo diode patch shape: 0 rectangle, 1 ellipse.

			====PhotoDiodePatchActiveColor====
			Photo diode patch color when active (RGB color in format 0xrrggbb).

			====PhotoDiodePatchInactiveColor====
			Photo diode patch color when inactive, (RGB color in format 0xrrggbb, use 0xff000000 for transparent).

			====PhotoDiodePatchExpression====
			Photo diode patch expression, determines active state by evaluating to nonnull. Defaults to <tt>StimulusBegin>0</tt>.

Mellinger: /* MinimumEvidence */

2013-04-30T15:21:44Z

MinimumEvidence

← Older revision		Revision as of 15:21, 30 April 2013
Line 50:		Line 50:

	====MinimumEvidence====		====MinimumEvidence====
			'''NOTE:''' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [[Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|these notes on how to use a different classifier]].

	By default, target selection is performed without considering the actual amount of evidence that favors the selected target over other targets. Although the selected target will always be a target with maximum classification score (i.e., evidence), other targets may have the same or a similar score. It may be useful to omit classification in such situations altogether, by specifying a minimum amount of evidence that must exist in favor of the selected target, when compared to the next-best target. When used together with the ''AccumulateEvidence'' option, this allows to dynamically control the number of stimulus presentations, by simply repeating stimulus sequences until a sufficient amount of evidence has been collected.		By default, target selection is performed without considering the actual amount of evidence that favors the selected target over other targets. Although the selected target will always be a target with maximum classification score (i.e., evidence), other targets may have the same or a similar score. It may be useful to omit classification in such situations altogether, by specifying a minimum amount of evidence that must exist in favor of the selected target, when compared to the next-best target. When used together with the ''AccumulateEvidence'' option, this allows to dynamically control the number of stimulus presentations, by simply repeating stimulus sequences until a sufficient amount of evidence has been collected.

Mellinger: /* AccumulateEvidence */

2013-04-30T15:20:05Z

AccumulateEvidence

← Older revision		Revision as of 15:20, 30 April 2013
Line 95:		Line 95:
	In addition, accumulated overall evidence will not increase if there is no consistent evidence in favor of a certain target. Thus, it is possible to operate a P300 BCI in a quasi-asynchronous mode by using ''AccumulateEvidence'', and choosing a ''MinimumEvidence'' value that is large enough to make accidental selection unlikely. In this configuration, no selection will be made unless the BCI user is actually concentrating on a target for a number of stimulus presentations, resulting in consistently accumulating evidence for that target.		In addition, accumulated overall evidence will not increase if there is no consistent evidence in favor of a certain target. Thus, it is possible to operate a P300 BCI in a quasi-asynchronous mode by using ''AccumulateEvidence'', and choosing a ''MinimumEvidence'' value that is large enough to make accidental selection unlikely. In this configuration, no selection will be made unless the BCI user is actually concentrating on a target for a number of stimulus presentations, resulting in consistently accumulating evidence for that target.

	''NOTE:'' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [{Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|these notes on ~~using~~ a different classifier]].		'''NOTE:''' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [[Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|these notes on how to use a different classifier]].

Mellinger: /* AccumulateEvidence */

2013-04-30T15:19:03Z

AccumulateEvidence

← Older revision		Revision as of 15:19, 30 April 2013
Line 95:		Line 95:
	In addition, accumulated overall evidence will not increase if there is no consistent evidence in favor of a certain target. Thus, it is possible to operate a P300 BCI in a quasi-asynchronous mode by using ''AccumulateEvidence'', and choosing a ''MinimumEvidence'' value that is large enough to make accidental selection unlikely. In this configuration, no selection will be made unless the BCI user is actually concentrating on a target for a number of stimulus presentations, resulting in consistently accumulating evidence for that target.		In addition, accumulated overall evidence will not increase if there is no consistent evidence in favor of a certain target. Thus, it is possible to operate a P300 BCI in a quasi-asynchronous mode by using ''AccumulateEvidence'', and choosing a ''MinimumEvidence'' value that is large enough to make accidental selection unlikely. In this configuration, no selection will be made unless the BCI user is actually concentrating on a target for a number of stimulus presentations, resulting in consistently accumulating evidence for that target.

	''NOTE:'' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|~~this information~~].		''NOTE:'' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [{Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|these notes on using a different classifier]].

Mellinger: /* AccumulateEvidence */

2013-04-30T15:18:06Z

AccumulateEvidence

← Older revision		Revision as of 15:18, 30 April 2013
Line 94:		Line 94:

	In addition, accumulated overall evidence will not increase if there is no consistent evidence in favor of a certain target. Thus, it is possible to operate a P300 BCI in a quasi-asynchronous mode by using ''AccumulateEvidence'', and choosing a ''MinimumEvidence'' value that is large enough to make accidental selection unlikely. In this configuration, no selection will be made unless the BCI user is actually concentrating on a target for a number of stimulus presentations, resulting in consistently accumulating evidence for that target.		In addition, accumulated overall evidence will not increase if there is no consistent evidence in favor of a certain target. Thus, it is possible to operate a P300 BCI in a quasi-asynchronous mode by using ''AccumulateEvidence'', and choosing a ''MinimumEvidence'' value that is large enough to make accidental selection unlikely. In this configuration, no selection will be made unless the BCI user is actually concentrating on a target for a number of stimulus presentations, resulting in consistently accumulating evidence for that target.

			''NOTE:'' If you are using your own classifier, this feature will not work properly unless your classifier's output matches certain criteria. Make sure to read [Programming_Reference:StimulusTask_Class#Using_a_Different_Classifier_with_StimulusTask\|this information].

Mellinger: /* MinimumEvidence */

2013-03-19T15:29:43Z

MinimumEvidence

← Older revision		Revision as of 15:29, 19 March 2013
Line 62:		Line 62:
	For large <math>\chi</math>, this relationship may be approximated and expressed in terms of error probability <math>p</math>:		For large <math>\chi</math>, this relationship may be approximated and expressed in terms of error probability <math>p</math>:

	<math>\eta \approx -\frac{\log_{10} p}{\log_{10} e} = -2.3\log_{10} p\, \textrm{for}\, p \leq 5%</math>.		<math>\eta \approx -\frac{\log_{10} p}{\log_{10} e} = -2.3\log_{10} p\quad\textrm{for}\; p \leq 5%</math>.

	Thus, the evidence value roughly corresponds to twice the number of leading zeros in the desired error probability, if classifier weights are normalized. Some values are provided in the following table:		Thus, the evidence value roughly corresponds to twice the number of leading zeros in the desired error probability, if classifier weights are normalized. Some values are provided in the following table:

Mellinger: /* MinimumEvidence */

2013-03-19T15:27:10Z

MinimumEvidence

← Older revision		Revision as of 15:27, 19 March 2013
Line 56:		Line 56:
	The meaning of the actual number entered into the ''MinimumEvidence'' parameter is relative to the amount of within-class variance present in the classification score. An evidence of 0 means a 50:50 chance for correct classification. Increasing the evidence value by two standard deviations corresponds to an improvement by a factor of roughly 88:12, four standard deviations correspond to (88:12)^2=(98:2) ... etc, approaching perfect classification as evidence increases towards infinity.		The meaning of the actual number entered into the ''MinimumEvidence'' parameter is relative to the amount of within-class variance present in the classification score. An evidence of 0 means a 50:50 chance for correct classification. Increasing the evidence value by two standard deviations corresponds to an improvement by a factor of roughly 88:12, four standard deviations correspond to (88:12)^2=(98:2) ... etc, approaching perfect classification as evidence increases towards infinity.

	~~P300GUI normalizes its~~ weights ~~to unit~~ within-class variance. In this case, you may use the following equations to convert between the ''MinimumEvidence'' parameter <math>\eta</math>, and the correct classification chance <math>\chi</math>:		In classifier training, classifier weights may be normalized such that within-class variance is 1 (this is done by recent versions of the P300Classifier tool). In this case, you may use the following equations to convert between the ''MinimumEvidence'' parameter <math>\eta</math>, and the correct classification chance <math>\chi</math>:

	<math>\chi = \frac{50}{50} \left(\frac{88}{12}\right)^{\eta/2}</math>		<math>\chi = \frac{50}{50} \left(\frac{88}{12}\right)^{\eta/2}</math>.

	<math>~~\eta=2\frac{\log~~\chi~~}{\log(88~~/~~12)}~~</math>.		For large <math>\chi</math>, this relationship may be approximated and expressed in terms of error probability <math>p</math>:

	~~For small error probabilities~~ <math>p < 5\%</math>~~, as required in practical applications, the approximations~~		<math>\eta \approx -\frac{\log_{10} p}{\log_{10} e} = -2.3\log_{10} p\, \textrm{for}\, p \leq 5%</math>.

	~~<math>-\ln(1-p) \approx p \ll \| \ln p \|</math>~~		Thus, the evidence value roughly corresponds to twice the number of leading zeros in the desired error probability, if classifier weights are normalized. Some values are provided in the following table:

	~~hold, which may be applied to obtain the very useful relation~~

	~~<math>\eta \approx -\frac{\log_{10} p}{\log_{10} e} = -2.3\log_{10} p</math>.~~

	Thus, the evidence value roughly corresponds to twice the number of leading zeros in the desired error probability. ~~Thus, we have~~ the following ~~correspondences in case of a normalized classifier~~:
	{\|		{\|
	!Selection Error		!Selection Error

Mellinger: /* MinimumEvidence */

2013-03-19T15:06:14Z

MinimumEvidence

← Older revision		Revision as of 15:06, 19 March 2013
Line 56:		Line 56:
	The meaning of the actual number entered into the ''MinimumEvidence'' parameter is relative to the amount of within-class variance present in the classification score. An evidence of 0 means a 50:50 chance for correct classification. Increasing the evidence value by two standard deviations corresponds to an improvement by a factor of roughly 88:12, four standard deviations correspond to (88:12)^2=(98:2) ... etc, approaching perfect classification as evidence increases towards infinity.		The meaning of the actual number entered into the ''MinimumEvidence'' parameter is relative to the amount of within-class variance present in the classification score. An evidence of 0 means a 50:50 chance for correct classification. Increasing the evidence value by two standard deviations corresponds to an improvement by a factor of roughly 88:12, four standard deviations correspond to (88:12)^2=(98:2) ... etc, approaching perfect classification as evidence increases towards infinity.

	~~You may use the ''Normalize Within-Class Variance'' option of P300Classifier in order to obtain classifier~~ weights ~~normalized~~ to unit within-class variance. In this case, you may use the following equations to convert between the ''MinimumEvidence'' parameter <math>\eta</math>, and the correct classification chance <math>\chi</math>:		P300GUI normalizes its weights to unit within-class variance. In this case, you may use the following equations to convert between the ''MinimumEvidence'' parameter <math>\eta</math>, and the correct classification chance <math>\chi</math>:

	<math>\chi = \frac{50}{50} \left(\frac{88}{12}\right)^{\eta/2}</math>		<math>\chi = \frac{50}{50} \left(\frac{88}{12}\right)^{\eta/2}</math>