UNIX and Linux SAS^® administrators, have you ever been greeted by one of these statements as you walk into the office before you have gotten your first cup of coffee? Power outage! SAS servers are down. I cannot access my reports. Have you frantically tried to restart the SAS servers to avoid loss of productivity and missed one of the steps in the process, causing further delays while other work continues to pile up? If you have had this experience, you understand the benefit to be gained from a utility that automates the management of these multi-tiered deployments. Until recently, there was no method for automatically starting and stopping multi-tiered services in an orchestrated fashion. Instead, you had to use time-consuming manual procedures to manage SAS services. These procedures were also prone to human error, which could result in corrupted services and additional time lost, debugging and resolving issues injected by this process. To address this challenge, SAS Technical Support created the SAS Local Services Management (SAS_lsm) utility, which provides automated, orderly management of your SAS^® multi-tiered deployments. The intent of this paper is to demonstrate the deployment and usage of the SAS_lsm utility. Now, go grab a coffee, and let's see how SAS_lsm can make life less chaotic.

Machine learning is in high demand. Whether you are a citizen data scientist who wants to work interactively or you are a hands-on data scientist who wants to code, you have access to the latest analytic techniques with SAS^® Visual Data Mining and Machine Learning on SAS^® Viya . This offering surfaces in-memory machine learning techniques such as gradient boosting, factorization machines, neural networks, and much more through its interactive visual interface, SAS^® Studio tasks, procedures, and a Python client. Learn about this multi-faceted new product and see it in action.

Machine learning predictive modeling algorithms are governed by hyperparameters that have no clear defaults agreeable to a wide range of applications. A few examples of quantities that must be prescribed for these algorithms are the depth of a decision tree, number of trees in a random forest, number of hidden layers and neurons in each layer in a neural network, and degree of regularization to prevent overfitting. Not only do ideal settings for the hyperparameters dictate the performance of the training process, but more importantly they govern the quality of the resulting predictive models. Recent efforts to move from a manual or random adjustment of these parameters have included rough grid search and intelligent numerical optimization strategies. This paper presents an automatic tuning implementation that uses SAS/OR^® local search optimization for tuning hyperparameters of modeling algorithms in SAS^® Visual Data Mining and Machine Learning. The AUTOTUNE statement in the NNET, TREESPLIT, FOREST, and GRADBOOST procedures defines tunable parameters, default ranges, user overrides, and validation schemes to avoid overfitting. Given the inherent expense of training numerous candidate models, the paper addresses efficient distributed and parallel paradigms for training and tuning in SAS^® Viya . It also presents sample tuning results that demonstrate improved model accuracy over default configurations and offers recommendations for efficient and effective model tuning.

Let's walk through an example of communicating from the SAS^® client to SAS^® Viya . The demonstration focuses on how to use SAS^® language to establish a session, transport and persist data, and receive results. Learn how to establish communication with SAS Viya. Explore topics such as: What is a session? How do I make requests? What does my SAS log tell me? Get a deeper understanding of data location on the client and the server side. Learn about applying existing user formats, how to get listings or reports, and how to query sessions, data, and properties.

SAS^® is perfect for building enterprise apps. Think about it: SAS speaks to almost any database you can think of and is probably already hooked in to most of the data sources in your organization. A full-fledged metadata security layer happens to already be integrated with your single sign-on authentication provider, and every time a user interacts with the system, their permissions are checked and the data their app asks for is automatically encrypted. SAS ticks all the boxes required by IT, and the skills required to start developing apps already sit within your department. Your team most likely already knows what your app needs to do, so instead of writing lists of requirements, give them an HTML5 resource, and together they can write and deploy the back-end code themselves. The apps run in the browser, the server-side logic is deployed using SAS .spk packages, and permissions are managed via SAS^® Management Console. Best of all, the infrastructure that would normally take months to integrate is already there, eliminating barriers to entry and letting you demonstrate the value of your solution to internal customers with zero up-front investment. This paper shows how SAS integrates with open-source tools like H54S, AngularJS, and PostGIS, together with next-generation developer-centric analytical platforms like SAS^® Viya , to build secure, enterprise-class apps that can support thousands of users. This presentation includes lots of app demos. This presentation was included at SAS^® Forum UK 2016.

The DATA step is the familiar and powerful data processing language in SAS^® and now SAS Viya . The DATA step's simple syntax provides row-at-a-time operations to edit, restructure, and combine data. New to the DATA step in SAS Viya are a varying-size character data type and parallel execution. Varying-size character data enables intuitive string operations that go beyond the 32KB limit of current DATA step operations. Parallel execution speeds the processing of big data by starting the DATA step on multiple machines and dividing data processing among threads on these machines. To avoid multi-threaded programming errors, the run-time environment for the DATA step is presented along with potential programming pitfalls. Come see how the DATA step in SAS Viya makes your data processing simpler and faster.

Hash objects have been supported in the DATA step and in the FCMP procedure for a while, but have you ever felt that hash objects could do a little more? For example, what if you needed to store more than doubles and character strings? Introducing PROC FCMP dictionaries. Dictionaries allow you to create references not only to numeric and character data, but they also give you fast in-memory hashing to arrays, other dictionaries, and even PROC FCMP hash objects. This paper gets you started using PROC FCMP dictionaries, describes usage syntax, and explores new programming patterns that are now available to your PROC FCMP programs, functions, and subroutines in the new SAS^® Viya platform environment.

Factorization machines are a new type of model that is well suited to very high-cardinality, sparsely observed transactional data. This paper presents the new FACTMAC procedure, which implements factorization machines in SAS^® Visual Data Mining and Machine Learning. This powerful and flexible model can be thought of as a low-rank approximation of a matrix or a tensor, and it can be efficiently estimated when most of the elements of that matrix or tensor are unknown. Thanks to a highly parallel stochastic gradient descent optimization solver, PROC FACTMAC can quickly handle data sets that contain tens of millions of rows. The paper includes examples that show you how to use PROC FACTMAC to recommend movies to users based on tens of millions of past ratings, predict whether fine food will be highly rated by connoisseurs, restore heavily damaged high-resolution images, and discover shot styles that best fit individual basketball players. ^®

In this course you will learn how to access and manage SAS and Excel data in SAS^® Viya .

The openness of SAS^® Viya , the new cloud analytic platform that uses SAS^® Cloud Analytic Services (CAS), emphasizes a unified experience for data scientists. You can now execute the analytics capabilities of SAS^® in different programming languages including Python, Java, and Lua, as well as use a RESTful endpoint to execute CAS actions directly. This paper provides an introduction to these programming languages. For each language, we illustrate how the API is surfaced from the CAS server, the types of data that you can upload to a CAS server, and the result tables that are returned. This paper also provides a comprehensive comparison of using these programming languages to build a common analytical process, including loading data to a CAS server; exploring, manipulating, and visualizing data; and building statistical and machine learning models.

What if you had analytics near the edge for your Internet of Things (IoT) devices that would tell you whether a piece of equipment is operating within its normal range? And what if those same analytics could help you intelligently determine what data you should keep and what data should be filtered at the edge? This session focuses on classifying streaming data near the edge by showcasing a demo that implements a single-class classification model within a gateway device. The model identifies observations that are normal and abnormal to help determine possible machine issues and preventative maintenance opportunities. The classification also helps to provide a method for filtering data at the edge by capturing all abnormal data but taking only a sample of the normal operating data. The model is developed using SAS^® Viya and implemented on a gateway device using SAS^® Event Stream Processing. By using a single-class classification technique, the demo also illustrates how to avoid issues with binary classification that would require failure observations in order to build an accurate model. Problems that this demo addresses include: identifying potential and future equipment failures in near real time; filtering sensor data near the edge to prevent unnecessary transport and storage of less valuable data; and building a classification model for failure that doesn't require observations relating to failures.

For many years now you have learned the ins and outs of using SAS/ACCESS^® software to move data into SAS^® to do your analytics. With the new open, cloud-ready SAS^® Viya platform comes a new set of data access technologies known as SAS data connectors and SAS data connect accelerators. This paper describes what these new data access products are and how they integrate with the SAS Viya platform. After reading this paper, you will have the foundation needed to load data from third-party data sources into SAS Viya.

Geofencing is one of the most promising and exciting concepts that has developed with the advent of the internet of things. Like John Anderton in the 2002 movie Minority Report, you can now enter a mall and immediately receive commercial ads and offers based on your personal taste and past purchases. Authorities can track vessels positions and detect when a ship is not in the area it should be, or they can forecast and optimize harbor arrivals. When a truck driver breaks from the route, the dispatcher can be alerted and can act immediately. And there are countless examples from manufacturing, industry, security, or even households. All of these applications are based on the core concept of geofencing, which consists of detecting whether a device s position is within a defined geographical boundary. Geofencing requires real-time processing in order to react appropriately. In this session, we explain how to implement real-time geofencing on streaming data with SAS^® Event Stream Processing and achieve high-performance processing, in terms of millions of events per second, over hundreds of millions of geofences.

Not only does the new SAS^® Viya platform bring exciting advancements in high-performance analytics, it also takes a revolutionary step forward in the area of administration. The new SAS^® Cloud Analytic Services is accompanied by new platform management tools and techniques that are designed to ease the administrative burden while leveraging the open programming and visual interfaces that are standard among SAS Viya applications. Learn about the completely rewritten SAS^® Environment Manager 3.2, which supports the SAS Viya platform. It includes a cleaner HTML5-based user interface, more flexible and intuitive authorization windows, and user and group management that is integrated with your corporate Lightweight Directory Access Protocol (LDAP). Understand how authentication works in SAS Viya without metadata identities. Discover the key differences between SAS^®9 and SAS Viya deployments, including installation and automated update-in-place strategies orchestrated by Ansible for hot fixes, maintenance, and new product versions alike. See how the new microservices and stateful servers are managed and monitored. In general, gain a better understanding of the components of the SAS Viya architecture, and how they can be collectively managed to keep your environment available, secure, and performant for the users and processes you support.

Ensemble models have become increasingly popular in boosting prediction accuracy over the last several years. Stacked ensemble techniques combine predictions from multiple machine learning algorithms and use these predictions as inputs to a second level-learning algorithm. This paper shows how you can generate a diverse set of models by various methods (such as neural networks, extreme gradient boosting, and matrix factorizations) and then combine them with popular stacking ensemble techniques, including hill-climbing, generalized linear models, gradient boosted decision trees, and neural nets, by using both the SAS^® 9.4 and SAS^® Visual Data Mining and Machine Learning environments. The paper analyzes the application of these techniques to real-life big data problems and demonstrates how using stacked ensembles produces greater prediction accuracy than individual models and na ve ensembling techniques. In addition to training a large number of models, model stacking requires the proper use of cross validation to avoid overfitting, which makes the process even more computationally expensive. The paper shows how to deal with the computational expense and efficiently manage an ensemble workflow by using parallel computation in a distributed framework.

SAS^® Cloud Analytic Services (CAS) is the cloud-based run-time environment for data management and analytics in SAS^®. By run-time environment, we refer to the combination of hardware and software where data management and analytics take place. In a sense, CAS is just another SAS platform to do things. CAS is a platform for high-performance analytics and distributed computing. The CAS server provides data management and an analytics framework that can run in the cloud, that can act as a cloud, and that provides the best-in-class analytics that SAS is known for. This new architecture functions as a public API, allowing access from many different clients such as Lua, Python, Java, REST, and yes, even SAS. The CAS server is designed to provide user-level sessions, to share data between sessions, and to provide fault tolerance, which allows a worker node to crash without losing data and allows the user action to continue running to completion. The isolation provided to each session allows one session to crash without affecting other sessions. The concept of 'always in memory' in CAS means that an action is not aware of what the server does to allow the action to access the data. The entire file might be in memory or just pieces of the file might be mapped into memory, just in time for the action to access the data. This allows CAS tables to be loaded that are larger than the memory available across the grid. Hadoop can be used to provide data redundancy. The server is elastic and can add or remove nodes as needed. Users can specify how many nodes they want their session to use, so that the session fits their needs.

Trends in predictive modeling, specifically in machine learning, are moving toward automated approaches where well-known predictive algorithms are applied and the best one is chosen according to some evaluation metric. This approach's efficacy relies on the underlying data preparation in general, and on data preprocessing in particular. Commonly used data preprocessing techniques include missing value imputation, outlier detection and treatment, functional transformation, discretization, and nominal grouping. Excluding toy problems, the composition of the best data preprocessing step depends on the modeling task at hand. This necessitates an iterative generation and evaluation of predictive pipelines, which consist of a mix of data preprocessing techniques and predictive algorithms. This is a combinatorial problem that can be a bottleneck in the analytics workflow. In this paper, we discuss the SAS^® Cloud Analytic Services (CAS) actions in SAS^® Viya that can be used to effect this end-to-end predictive pipeline in a scalable way, with special emphasis on CAS actions for data exploration, preprocessing, and feature transformation. In addition, we discuss how the whole process can be automated.

Transport Layer Security (TLS) configuration for SAS^® components is essential to protect data in motion. All necessary encryption arrangement is established through a TLS handshake between the client and the server side. Many SAS^® 9.4 and SAS^® Viya components can be a client side, a server side, or both. SAS documentation primarily provides how-to steps for the configuration. This paper examines the X.509 certificate and the TLS handshake protocol, which are the basic building blocks of the secure communication. The paper focuses on the logic behind the setup and how various types of certificates are used in the configuration. Many unique client and server combinations of SAS components are illustrated and explained with the best-practice suggestions.

With SAS^® Viya and SAS^® Cloud Analytic Services (CAS), SAS is moving into a new territory where SAS^® Analytics is accessible to popular scripting languages using open APIs. Python is one of those client languages. We demonstrate how to connect to CAS, run CAS actions, explore data, build analytical models, and then manipulate and visualize the results using standard Python packages such as Pandas and Matplotlib. We cover a wide variety of topics to give you a bird's eye view of what is possible when you combine the best of SAS with the best of open source.