White Papers

Solving the World’s Unsolved Problems
About this paper

Response or propensity models are widely used by businesses to identify customers who are likely to purchase a product. To help develop these models, techniques like logistic regression and decision trees have been deployed across verticals such as banking, telecom, and retail. Advanced algorithms, such as random forest, gradient boosting techniques, and neural network models are also used for classifying outcomes, such as responder vs non-responder and good vs bad. Since propensity models are extremely important to a business, a lot of attention is paid to statistical validation of the selected model before it is deployed. Most importantly, even after the model has been validated and found to be robust, periodic monitoring is imperative to ensure that the model is performing at peak efficiency over a course of time. Ongoing monitoring is also required to determine whether changes in market conditions or business strategies demand adjustment, redevelopment, or replacement of the model.

Several metrics are used to monitor the model’s performance and its validity. This paper discusses the different ways of confirming the model’s validity. In addition, it attempts to answer the question of when to retire a model….

Read More
About this paper

Businesses use optimization techniques in decision-making at the tactical and strategic levels. Historically, it has been used to solve complex problems in areas such as production and distribution planning, inventory management, raw material sourcing, job scheduling, skill-based work allocation, and budget allocation….

Read More
About this paper

With vast amounts of data, the infrastructure demands for advanced analytics is ever increasing. There are a class of analytic problems like optimization, forecasting, machine learning in domains of retail, finance, and other verticals, where the problem could be broken down into smaller sub-problems. Example, in retail; assortment by store, or price optimization by category, in finance; security portfolio optimization by industry, the entire problem is broken down into a subset of smaller jobs. In this paper, we present orchestrating these jobs on Kubernetes, GKE powered by Google …

Read More
About this paper

Data-driven agent allocation provides an immense opportunity in improving the efficiency of any process. Cost effective system can be designed using machine learning and optimization in SAS®. In this paper, the application of machine learning and optimization for agent allocation in two-stage job processing is described using a case study of a Business Process Outsourcing (BPO) organization. This organization handles verification and underwriting processes for credit card applications of large banks. The paper provides a brief overview of unsupervised machine learning algorithms for agent …

Read More
About this paper

In this paper, we discuss a use-case wherein every quarter a business wants to generate more than 1500 customized customer reports, so we have developed a tool which enables the business to generate the customer specific reports on a single-click and share the report(s) with the customer(s) over an email attachment. To achieve this, we use SAS Programming, SAS Add-In for Microsoft Office (MS-Office), VB Script and Think-Cell plugin for MS-Office. The final output is in the form of a MS-PowerPoint and PDF report which can be shared with …

Read More
About this paper

One of the challenges business users talk to us about is, handling data of different variety and size and having the capacity to process and extract information from the data to make a business decision(s). SAS® Enterprise Guide (EG) can handle complex data files effortlessly with different file formats coming from various data sources, in addition, SAS Visual Analytics (VA) with its vibrant visuals can be used to develop business-friendly dashboards. Did you ever have a need to publish or update the comments and insights while …

Read More
About this paper

SAS Grid is a shared, centrally managed analytics computing environment that features workload balancing and management, high availability, and fast processing. A SAS Grid environment helps you incrementally scale your computing infrastructure over time as the number of users and the size of data grow. It also provides rolling maintenance and upgrades without any disruption to your users. This Quick Start bootstraps the infrastructure for a SAS Grid cluster and installs the SAS Grid software, which includes SAS Grid Control Server, SAS Grid nodes, SAS Metadata Server, …

Read More
About this paper

The SAS® Viya® platform comes with a new command line interface to interact with microservices. This paper is an attempt to embrace the openness of the Viya Platform by creating a chatbot, which helps SAS Administrator in performing his/her day to day tasks. While there are many ways to automate the Admin tasks, this paper explores the latest cloud services such as AWS Lex chatbot service, AWS Lambda which is a serverless computing platform to create an Interactive user chatbot with the front end being the Slack application …

Read More
About this paper

About this paper SAS® Analytics for Containers provides the option to deploy SAS® Analytics within a container- enabled infrastructures, including Docker and Kubernetes, which are often run in the cloud. With an aim to analyze massively large data from Google BigQuery through SAS® in a containerized environment, we have integrated Google BigQuery with SAS® 9.4 Analytics Pro in Docker Container on Google Cloud Environment. This paper guides you through the process of configuring SAS® Access to BigQuery in containerized SAS® Application and validation steps for …

Read More
About this paper

This paper talks about the orchestration of SAS® data integration processes based on the arrival of the SAS data integration input files in Amazon Web Services (AWS) S3. Our client runs a daily process where they generate credit statements for their customers. Each customer receives their statement once in a month. Every day, around 200K customers are processed, eventually reaching out to their entire customer base of roughly 6 million in a month. The process starts in their on-premises datacenter, followed by the APR calculations in SAS data integration …

Read More
About this paper

Data-driven agent allocation provides an immense opportunity in improving the efficiency of any process. Cost effective system can be designed using machine learning and optimization in SAS®. In this paper, the application of machine learning and optimization for agent allocation in two-stage job processing is described using a case study of a Business Process Outsourcing (BPO) organization. This organization handles verification and underwriting processes for credit card applications of a large bank. The paper provides a brief overview of unsupervised machine learning algorithms for agent and application profiling. A Cluster of application and agents are created and are used in the agent allocation optimization problem. Optimization model framework is extensively discussed to solve the problem of skill-based agent allocation for credit card application processing based on its complexities, which has two stages of processing, application verification, and underwriting. A mixed integer optimization problem is modeled and is solved using the OPTMODEL procedure. The design for the end-to-end process to implement optimization for agent allocation is also discussed in this paper…

Read More