Deep Learning

What it is and why it matters

Deep learning in today's world

The impact of deep learning is significant – and it’s only getting started.

Deep learning and GANs: How they relate

A generative adversarial network (GAN) is a type of machine learning algorithm. GANs help data scientists create synthetic data for data-hungry deep learning models. That’s important because using synthetic data allows for the creation of deep learning models not previously possible.

Learn more in the blog post

Discover how SAS uses deep learning

This step-by-step guide compares multiple neural network models and explains how to use them. You'll get an introduction to deep learning techniques and applications and learn how SAS supports the creation of deep neural network models.

Download the paper

How to use deep learning for embedding images

Embedding models reduce the dimensionality of input data, such as images. With an embedding model, input images are converted into low-dimensional vectors – so it's easier for other computer vision tasks to use. The key is to train the model so similar images are converted to similar vectors.

Watch the video

Who’s using deep learning?

To the outsider, deep learning may appear to be in a research-and-development phase as computer science researchers and data scientists continue to test its capabilities. However, deep learning has many practical applications that businesses are using today, and many more that will be used as research continues. Popular uses today include:


In the retail industry, it’s important to stay one step ahead of customer expectations. Deep learning is making that happen. Using customer data, along with speech recognition and natural language processing, retailers can predict customer preferences and needs and cut down on needless stock. They can also assist in locating the best quality product for the lowest price. Ultimately, this helps retailers cater to their customers and saves retailers and customers time and money.


Deep learning enables banks to identify patterns in unstructured data and improve enterprisewide decision making. In risk management, deep learning helps banks interrogate multiple sources and set appropriate lending limits with greater confidence. All without compromising fairness. Deep learning also plays a vital role in detecting and preventing fraud and financial crime. One example is monitoring real-time video, identifying suspicious activity in branches or ATM locations, and preventing an account takeover if the voice biometric profile does not match the genuine customer. And when it comes to delivering an improved user experience, deep learning's ability to assist with sentiment analysis ensures issues reported via social channels can be swiftly identified and remedied.


Manufacturing uses deep learning and other AI techniques to improve the overall quality of the industry. One of the largest expenses in this industry is the maintenance of equipment, and deep learning is instrumental in reducing or avoiding downtime of crucial resources and equipment. The use of deep learning and capabilities, like computer vision, identifies quality problems using object detection, process monitoring and anomaly detection. The manufacturing industry can save money from unplanned downtime, better-designed products, improved efficiency, product quality and overall employee safety.

Health Care

Deep learning supports the health care industry by ensuring better patient care and operational efficiency. With deep learning, health care professionals can analyze data faster and more precisely. Electronic health records can be created faster and with fewer errors using speech-to-text with natural language processing tools. Neural networks, paired with image recognition, analyze medical images instead of just reading them, helping health care professionals identify tumors and their progression.

Transportation & Logistics

Deep learning can help the travel and logistics industry increase productivity and efficient operational planning. Using predicative software, the industry can stay ahead of potential failures and up to date on scheduled truck repairs to reduce operating costs. Along with predictive maintenance, deep learning and AI can track vehicles in real time, allowing hauling companies to locate and monitor the speed of their fleet in real time. This is all made possible with the precision and speed of deep learning.


Government agencies can use deep learning to improve fraud detection using handwriting analysis and land and water management using image recognition. Deep learning also helps create a better understanding of citizen preferences through natural language translation of sentiment. Another example is reducing infrastructure spending by using predictive maintenance capabilities. Overall, deep learning allows governments to solve problems that were too challenging to address previously.


Deep learning helps more value from the many data types in call center operations. Other ways deep learning support utilities include recommending specific corrective actions for line and equipment maintenance, vegetation management and a myriad of forecasting functions (from sales forecasts and net load forecasts to load forecasts and more). Indeed, deep learning will be a foundation part of a utility’s position in the future.

How deep learning works 

Deep learning changes how you think about representing the problems that you’re solving with analytics. It moves from telling the computer how to solve a problem to training the computer to solve the problem itself.


Feature representation

Deep learning is a paradigm shift in model building that moves from feature engineering to feature representation. 

Deep learning layers

Instead of using known variables to predict unknowns, deep learning uses looks at layers of the data to recognize latent features of the data. 

Deep learning results

The promise of deep learning is that it can lead to predictive systems that generalize well, adapt well, continuously improve as new data arrives. You no longer fit a model. Instead, you train the task.

A traditional approach to analytics is to use the data at hand to engineer features to derive new variables, then select an analytic model and finally estimate the parameters (or the unknowns) of that model. These techniques can yield predictive systems that do not generalize well because completeness and correctness depend on the quality of the model and its features. For example, if you develop a fraud model with feature engineering, you start with a set of variables, and you most likely derive a model from those variables using data transformations. You may end up with 30,000 variables that your model depends on, then you have to shape the model, figure out which variables are meaningful, which ones are not, and so on. Adding more data requires you to do it all over again.

The new approach with deep learning is to replace the formulation and specification of the model with hierarchical characterizations (or layers) that learn to recognize latent features of the data from the regularities in the layers.

SAS ile bağlantı kurun ve sizin için neler yapabileceğimizi görün.