Overfitting Challenge - Revision history

BPeat at 00:13, 20 September 2020

2020-09-20T00:13:38Z

BPeat at 00:10, 20 September 2020

2020-09-20T00:10:43Z

BPeat at 00:09, 20 September 2020

2020-09-20T00:09:18Z

BPeat at 04:57, 19 September 2020

2020-09-19T04:57:09Z

BPeat at 12:22, 21 August 2020

2020-08-21T12:22:58Z

BPeat at 00:56, 9 August 2020

2020-08-09T00:56:24Z

BPeat at 00:54, 9 August 2020

2020-08-09T00:54:07Z

BPeat at 11:54, 2 June 2020

2020-06-02T11:54:41Z

BPeat at 13:46, 23 February 2020

2020-02-23T13:46:18Z

BPeat at 05:43, 13 July 2019

2019-07-13T05:43:48Z

@@ Line 31: / Line 31: @@
 ** [[L1 and L2 Regularization]] -  update the general cost function by adding another term known as the regularization term.
 ** [[Dropout]] - at every iteration, it randomly selects some nodes and temporarily removes the nodes (along with all of their incoming and outgoing connections)
-** [[Data Augmentation, Data Labeling, and Auto-Tagging|Data Augmentation]]
+** [[Data Quality#Data Augmentation, Data Labeling, and Auto-Tagging|Data Augmentation]]
 ** [[Early Stopping]]

@@ Line 25: / Line 25: @@
 * add more data
-* use [[Batch Norm(alization) & Standardization]]
+* use [[Data Quality#Batch Norm(alization) & Standardization|Batch Norm(alization) & Standardization]]
 * use architectures that generalize well
 * reduce architecture complexity

@@ Line 25: / Line 25: @@
 * add more data
 * use [[Batch Norm(alization) & Standardization]]
 * use architectures that generalize well
@@ Line 32: / Line 31: @@
 ** [[L1 and L2 Regularization]] -  update the general cost function by adding another term known as the regularization term.
 ** [[Dropout]] - at every iteration, it randomly selects some nodes and temporarily removes the nodes (along with all of their incoming and outgoing connections)
-** [[Data Augmentation]]
+** [[Data Augmentation, Data Labeling, and Auto-Tagging|Data Augmentation]]
 ** [[Early Stopping]]

@@ Line 25: / Line 25: @@
 * add more data
-* use [[Data Augmentation]]
+* use [[Data Augmentation, Data Labeling, and Auto-Tagging|Data Augmentation]]
 * use [[Batch Norm(alization) & Standardization]]
 * use architectures that generalize well

@@ Line 10: / Line 10: @@
 * [http://www.unite.ai/what-is-overfitting/ What Is Overfitting? | Daniel Nelson - Unite.ai]
 * [http://docs.aws.amazon.com/machine-learning/latest/dg/evaluating_models.html#overfitting Preventing Overfitting | AWS]
-* [[Approach to Bias and Variances]]
+* [[Bias and Variances]]
 Overfitting is a problem in machine learning in general, not just in neural networks. The problem is inherent in the way machine learning models are developed: A set of "training data" is used to "train" the model. The goal is to have a model that can then be used on data that hasn't been seen before. Over-fitting refers to the problem of having the model trained to work so well on the training data that it starts to work more poorly on data it hasn't seen before. There are a number of techniques to mitigate or prevent over-fitting. [http://wiki.fast.ai/index.php/Over-fitting | Deep Learning Course Wiki]

@@ Line 37: / Line 37: @@
 http://s3-ap-south-1.amazonaws.com/av-blog-media/wp-content/uploads/2018/04/Screen-Shot-2018-04-04-at-2.43.37-PM-768x592.png
 <youtube>ms-Ooh9mjiE</youtube>
 <youtube>-JopeGg60QY</youtube>

← Older revision		Revision as of 11:54, 2 June 2020
Line 9:		Line 9:

	* [http://www.unite.ai/what-is-overfitting/ What Is Overfitting? \| Daniel Nelson - Unite.ai]		* [http://www.unite.ai/what-is-overfitting/ What Is Overfitting? \| Daniel Nelson - Unite.ai]
		+	* [http://docs.aws.amazon.com/machine-learning/latest/dg/evaluating_models.html#overfitting Preventing Overfitting \| AWS]

	Overfitting is a problem in machine learning in general, not just in neural networks. The problem is inherent in the way machine learning models are developed: A set of "training data" is used to "train" the model. The goal is to have a model that can then be used on data that hasn't been seen before. Over-fitting refers to the problem of having the model trained to work so well on the training data that it starts to work more poorly on data it hasn't seen before. There are a number of techniques to mitigate or prevent over-fitting. [http://wiki.fast.ai/index.php/Over-fitting \| Deep Learning Course Wiki]		Overfitting is a problem in machine learning in general, not just in neural networks. The problem is inherent in the way machine learning models are developed: A set of "training data" is used to "train" the model. The goal is to have a model that can then be used on data that hasn't been seen before. Over-fitting refers to the problem of having the model trained to work so well on the training data that it starts to work more poorly on data it hasn't seen before. There are a number of techniques to mitigate or prevent over-fitting. [http://wiki.fast.ai/index.php/Over-fitting \| Deep Learning Course Wiki]

@@ Line 7: / Line 7: @@
 [http://www.youtube.com/results?search_query=Regularization+Dropout+Overfitting Youtube search...]
 [http://www.google.com/search?q=Regularization+Dropout+deep+machine+learning+ML ...Google search]
 Overfitting is a problem in machine learning in general, not just in neural networks. The problem is inherent in the way machine learning models are developed: A set of "training data" is used to "train" the model. The goal is to have a model that can then be used on data that hasn't been seen before. Over-fitting refers to the problem of having the model trained to work so well on the training data that it starts to work more poorly on data it hasn't seen before. There are a number of techniques to mitigate or prevent over-fitting. [http://wiki.fast.ai/index.php/Over-fitting | Deep Learning Course Wiki]

← Older revision		Revision as of 05:43, 13 July 2019
Line 14:		Line 14:
	Trade-off is tension between the error introduced by the bias and the variance. [http://machinelearningmastery.com/gentle-introduction-to-the-bias-variance-trade-off-in-machine-learning/ Jason Brownlee]		Trade-off is tension between the error introduced by the bias and the variance. [http://machinelearningmastery.com/gentle-introduction-to-the-bias-variance-trade-off-in-machine-learning/ Jason Brownlee]

		+	http://miro.medium.com/max/468/1*xwtSpR_zg7j7zusa4IDHNQ.png
		+
		+	Whenever we discuss model prediction, it’s important to understand prediction errors (bias and variance). There is a tradeoff between a model’s ability to minimize bias and variance... To build a good model, we need to find a good balance between bias and variance such that it minimizes the total error.[http://towardsdatascience.com/understanding-the-bias-variance-tradeoff-165e6942b229 Understanding the Bias-Variance Tradeoff \| Seema Singh]

	Good practices for addressing overfitting:		Good practices for addressing overfitting: