保留正文与补采附件。源页面当前未重新核验；具体限制见交付清单。

Revisit consent button We value your privacy We use cookies to enhance your browsing experience, serve personalised ads or content, and analyse our traffic. By clicking "Accept All", you consent to our use of cookies. Customise Reject All Accept All Customise Consent Preferences We use cookies to help you navigate efficiently and perform certain functions. You will find detailed information about all cookies under each consent category below. The cookies that are categorised as "Necessary" are stored on your browser as they are essential for enabling the basic functionalities of the site. ... Show more NecessaryAlways Active Necessary cookies are required to enable the basic features of this site, such as providing secure log-in or adjusting your consent preferences. These cookies do not store any personally identifiable data.

Functional Functional cookies help perform certain functionalities like sharing the content of the website on social media platforms, collecting feedback, and other third-party features.

Analytics Analytical cookies are used to understand how visitors interact with the website. These cookies help provide information on metrics such as the number of visitors, bounce rate, traffic source, etc.

Performance Performance cookies are used to understand and analyse the key performance indexes of the website which helps in delivering a better user experience for the visitors.

Advertisement Advertisement cookies are used to provide visitors with customised advertisements based on the pages you visited previously and to analyse the effectiveness of the ad campaigns.

Uncategorised Other uncategorised cookies are those that are being analysed and have not been classified into a category as yet. No cookies to display. Reject All Save My Preferences Accept All Toggle SearchToggle Mobile Navigation Publish AI, ML & data-science insights to a global community of data professionals. Sign inSubmit an Article LatestEditor's PicksNewsletterShipAI

Write For TDS LinkedInXToggle Search Data Science

Avoid Overfitting in Neural Networks: a Deep Dive

Learn how to implement regularization techniques to boost performances and prevent Neural Network overfitting. Riccardo Andreoni November 30, 20239 min read

Learn how to implement regularization techniques to boost performances and prevent Neural Network overfitting

Image source: unsplash.com. Image source: unsplash.com. When training a deep neural network, it's often troublesome to achieve the same performances on both the training and validation sets. A considerably higher error on the validation set is a clear flag for overfitting : the network has become too specialized in the training data. In this article, I provide a comprehensive guide on how to bypass this issue. Learn this step by step with the interactive AI and Data Scientist, Computer Science and Machine Learning roadmaps.

Overfitting Neural Networks

When dealing with any machine learning application, it's important to have a clear understanding of the bias and variance of the model. In traditional machine learning algorithms, we talk about the bias vs. variance tradeoff, which consists of the struggle of minimizing both the variance and the bias of a model. In order to reduce the bias of a model (i.e. its error from erroneous assumptions), we need a more complex model. On the contrary, reducing the model's variance (the sensitivity of the model in capturing the variations of the training data), implies a more simple model. It is straightforward that the bias vs. variance tradeoff, in traditional machine learning, derives from the conflict of necessitating both a more complex and a simpler model at the same time. In the Deep Learning era, we have tools to reduce just the model's variance without hurting the model's bias or, on the contrary, to reduce the bias without increasing the variance. Before exploring the different techniques used to prevent the overfitting of a neural network, it's important to clarify what high variance or high bias means. Consider a common neural network task such as image recognition, and think over a neural network that recognizes the presence of pandas in a picture. We can confidently assess that a human can carry out this task with a near 0% error. As a consequence, this is a reasonable benchmark for the accuracy of the image recognition network. After training the neural network on the training set and evaluating its performances on both the training and validation sets, we may come up with these different results:

  1. Train Error = 20% and Validation Error = 22%
  2. Train Error = 1% and Validation Error = 15%
  3. Train Error = 0.5% and Validation Error = 1%
  4. Train Error = 20% and Validation Error = 30%

The first example is a typical instance of high bias : the error is large on both the train and validation sets. Conversely, the second example suffers from high variance , having a much lower accuracy when dealing with data the model didn't learn from. The third result represents low variance and bias , the model can be considered valid. Finally, the fourth example shows a case of both high bias and variance : not only the training error is large when compared with the benchmark, but the validation error is also higher. Image by the author. Image by the author. From now on, I will present several techniques of regularization, used to reduce the model's overfitting to the training data. They are beneficial for cases 2. and 4. of the previous example.

L1 and L2 Regularization

Similar to classical regression algorithms (linear, logistic, polynomial, etc.), L1 and L2 regularizations are also employed to prevent overfitting in high-variance neural networks. In order to keep this article short and on-point, I won't recall how L1 and L2 regularizations work on regression algorithms, but you can check this article for additional information. The idea behind the L1 and L2 regularization techniques is to constrain the model's weights to be smaller or to shrink some of them to 0. Consider the cost function J of a classical deep neural network: The cost function J is, of course, a function of the weights and biases of each layer 1, ..., L. m is the number of training examples and ℒ is the loss function.

L1 Regularization

In L1 Regularization we add the following term to the cost function J: where the matrix norm is the sum of the absolute value of the weights for each layer 1, ..., L of the network: λ is the regularization term. It's a hyperparameter that must be carefully tuned. λ directly controls the impact of the regularization: as λ increases, the effects on the weights shrinking are more severe. The complete cost function under L1 Regularization becomes: For λ=0, the effects of L1 Regularization are null. Instead, choosing a value of λ which is too big, will over-simplify the model, probably resulting in an underfitting network. L1 Regularization can be considered as a sort of neuron selection process because it would bring to zero the weights of some hidden neurons.

L2 Regularization

In L2 Regularization, the term we add to the cost function is the following: In this case, the regularization term is the squared norm of the weights of each network's layer. This matrix norm is called Frobenius norm and, explicitly, it's computed as follows: Please note that the weight matrix relative to layer l has n^{[l]} rows and n^{[l-1]} columns. Finally, the complete cost function under L2 Regularization becomes: Again, λ is the regularization term and for λ=0 the effects of L2 Regularization are null. L2 Regularization brings towards zero the values of the weights, resulting in a more simple model.

How do L1 and L2 Regularization reduce overfitting?

L1 and L2 Regularization techniques have positive effects on overfitting to the training data for two reasons:

The second reason is not trivial and deserves an expansion. Consider a hyperbolic tangent (tanh) activation function, whose graph is the following: Image by the author. Image by the author. From the function plot we can see that if the input value x is small, the function tanh(x) behaves almost linearly. When tanh is used as the activation function of a neural network's hidden layers, the input value is: which for small weights w is also close to zero. If each layer of the neural network is linear, we can prove that the whole network behaves linearly. Thus, constraining some of the hidden units to mimic linear functions, leads to a simpler network and, as a consequence, helps to prevent overfitting. A simpler model often is not capable to capture the noise in the training data and therefore, overfitting is less frequent.

Dropout

The idea of dropout regularization is to randomly remove some nodes in the network. Before the training process, we set a probability (suppose p = 50%) for each node of the network. During the training phase, each node has a p probability of being turned off. The dropout process is random, and it is performed separately for each training example. As a consequence, each training example might be trained on a different network. As for L2 Regularization, the result of dropout regularization is a simpler network, and a simpler network leads to a less complex model. The effect of dropout on a simple network. Image by the author. The effect of dropout on a simple network. Image by the author.

Dropout in practice

In this brief section, I show how to implement Dropout regularization in practice. I will go through a few simple lines of code (python). If you are only interested in the general theory of regularization, you can easily skip this section. Suppose we have stored the activation values of the network's layer 4 in the NumPy array a4. First, we create the auxiliary vector d4: python

# Set the keeping probability
keep_prob = 0.7

python

# Create auxiliary vector for layer 4
d4 = np.random.rand(a4.shape[0], a3.shape[1]) < keep_prob

The vector d4 has the same dimensions as a4 and contains the values True or False based on the probability keep_prob. If we set a keeping probability of 70%, that is the probability that a given hidden unit is kept, and therefore, the probability of having True value on a given element of d4. We apply the auxiliary vector d4 to the activations a4: python

# Apply the auxiliary vector d4 to the activation a4
a4 = np.multiply(a4,d4)

Finally, we need to scale the modified vector a4 by the keep_prob value: python

# Scale the modified vector a4
a4 /= keep_prob

This last operation is needed to compensate for the reduction of units in the layer. Carrying out this operation during the training process allows us not to apply dropout during the test phase.

How does dropout reduce overfitting?

Dropout has the effect of temporarily transforming the network into a smaller one , and we know that smaller networks are less complex and less prone to overfitting. Consider the network illustrated above, and focus on the first unit of the second layer. Because some of its inputs may be temporarily shut down due to dropout, the unit can't always rely on them during the training phase. As a consequence, the hidden unit is encouraged to spread its weights across its inputs. Spreading the weights has the effect of decreasing the squared norm of the weight matrix, resulting in a sort of L2 regularization. Setting the keeping probability is a fundamental step for an effective dropout regularization. Typically, the keeping probability is set separately for each layer of the neural network. For layers with a large weight matrix, we usually set a smaller keeping probability because, at each step, we want to conserve proportionally fewer weights with respect to smaller layers.

Other Regularization Techniques

In addition to L1/L2 regularization and dropout, there exist other regularization techniques. Two of them are data augmentation and early stopping. From the theory, we know that training a network on more data has positive effects on reducing high variance. As getting more data is often a demanding task, data augmentation is a technique that, for some applications, allows machine learning practitioners to get more data almost for free. In computer vision, data augmentation provides a larger training set by flipping, zooming, and translating the original images. In the case of digit recognition, we can also impose distortion on the images. Early stopping , as the name suggests, involves stopping the training phase before the initially defined number of iterations. If we plot the cost function on both the training set and the validation set as a function of the iterations, we can experience that, for an overfitting model, the training error always keeps decreasing but the validation error might start to increase after a certain number of iterations. When the validation error stops decreasing, that is exactly the time to stop the training process. By stopping the training process earlier, we force the model to be simpler, thus reducing overfitting.

Conclusion

In conclusion, the L1 and L2 regularization techniques are indispensable tools in addressing overfitting within neural networks. Image by the author. Image by the author. L1 regularization, with its ability to induce sparsity by penalizing irrelevant features , proves effective in simplifying models. On the other hand, L2 regularization, by penalizing the squared magnitude of weights, promotes smoother models , reducing the risk of extreme values. Dropout, a dynamic technique involving randomly deactivating neurons during the training phase, enhances model generalization. It prevents over-reliance on specific neurons, promoting a more robust network. However, these techniques come with tradeoffs. While regularization methods effectively mitigate overfitting, they may inadvertently restrict the model's ability to capture intricate patterns in the data. The real challenge is choosing the right balance between regularization strength and model complexity. For further insights, I will fill the Reference section with amazing resources. I highly suggest checking them out to refine your understanding of the subject.

If you liked this story, consider following me to be notified of my upcoming projects and articles! Here are some of my past projects:

Advanced Dimensionality Reduction Models Made Simple Use Deep Learning to Generate Fantasy Names: Build a Language Model from Scratch Outlier Detection with Scikit-Learn and Matplotlib: a Practical Guide

References

Register Now to Continue Reading

Unlock Towards Data Science with a Free Account

All articles on Towards Data Science are available to registered users. Create a free account today to access this article and thousands more. Register for freeSign in to your account

Written By

Riccardo Andreoni

See all from Riccardo Andreoni Data ScienceMachine LearningNeural NetworksOverfittingProgramming Share This Article

Towards Data Science is a community publication. Submit your insights to reach our global audience and earn through the TDS Author Payment Program. Write for TDS

Related Articles

Understand Support Vector Machines

Machine Learning part 1 of a 5-sequence, brief articles that provide a comprehensive introduction; no prior knowledge assumed Ahmar Shah, PhD (Oxford) March 7, 20216 min read

Latest picks: An Intuitive Look at the Basics of Information Theory

Machine Learning Your daily dose of data science TDS Editors January 19, 20211 min read

I Built a C++ Backend So My GPU Would Stop Eating Air

Machine Learning How to eliminate padding overhead and accelerate LLM inference at the hardware level. Anubhab Banerjee June 3, 202627 min read

Best places for new businesses in Florianopolis, Brazil (North Shore): a Foursquare data analysis

Data Science Which opportunities could the data show? Shaiala Marques September 14, 20209 min read

Graph Machine Learning @ ICML 2023

Artificial Intelligence Recent advancements and hot trends, August 2023 edition Michael Galkin August 6, 202315 min read

Classification With Rosenblatt's Perceptron

Machine Learning The "hello-world" of machine learning Pan Cretan September 8, 20237 min read

Sigmoid and SoftMax Functions in 5 minutes

Machine Learning The math behind two of the most used activation functions in Machine Learning Gabriel Furnieles September 8, 20225 min read

3 Reasons Why Spark's Lazy Evaluation is Useful

Programming Spark's Lazy Evaluation and Why You Should Care Amal Hasni September 14, 20225 min read

Bertelsman Arvato Financial Solution Customer Segmentation

Data Science Nowadays, with big data becomes reality, people now focus on how to use the data to realize commercial values. One area which is much more... Annie Wang January 28, 20218 min read

Your home for data science and AI. The world's leading publication for data science, data analytics, data engineering, machine learning, and artificial intelligence professionals. Subscribe to Our NewsletterWrite For TDS•About•Advertise•Privacy Policy•Terms of Use © Insight Media Group, LLC 2026•The New Stack•roadmap.sh•Creator Network

Recommended Articles

Close