To adhere to these laws, insurers either built or outsourced processes to match their policyholder data to the DMF or similar databases; however, the match criteria outlined in state regulatory agreements necessitated a secondary death validation to confirm a policyholder death – often a manual and time-consuming process. To complicate matters, thousands of erroneous deaths are reported in the DMF each year, further necessitating a sound validation process that mitigates the risk of a falsely reported death from a single source. A death validation process that cross-references deaths reported in the DMF against other death sources, such as state vital statistics and obituaries, can reduce or eliminate the manual process that insurance companies use to validate deaths. This streamlined process can be automated for millions of records and completed in a fraction of the time that it would require of a person or team. It should use common data elements available across multiple death sources, such as name and death date, to identify the same decedent, but also accommodate for slight variations (e.g., nicknames, misspellings) and false positive reduction (e.g., common names) that might otherwise result in no matches or mismatches.

MassMutual, a $30 billion per year life insurance company, had a problem. It was 2013 and, along with the rest of the insurance industry, it was bedeviled by fraud. According to FBI estimates, fraud sets the U.S. insurance industry (and policyholders) back by $40 billion a year. "We had to get much better at detecting fraud in real time," says Sears Merritt, MassMutual's chief of technology strategy and data science. So MassMutual launched an innovative collaboration between the company's data scientists and its line managers.

The goal of this case study is to explore storm events in various locations in the United States and analyze the frequency and damage costs associated with different types of events. A machine learning model is used to predict the damage costs, based on historical data from 1980 - 2018. The calculations are then performed in an app, which can be shared as a web application. This example also highlights techniques for preprocessing data in various forms (numeric, text, categorical, dates and times) and working with large data sets which do not fit into memory. The example is used in the "Data Science with MATLAB" webinar series.

DeepMind today announced a new milestone for its artificial intelligence agents trained to play the Blizzard Entertainment game StarCraft II. The Google-owned AI lab's more sophisticated software, still called AlphaStar, is now grandmaster level in the real-time strategy game, capable of besting 99.8 percent of all human players in competition. The findings are to be published in a research paper in the scientific journal Nature. Not only that, but DeepMind says it also evened the playing field when testing the new and improved AlphaStar against human opponents who opted into online competitions this past summer. For one, it trained AlphaStar to use all three of the game's playable races, adding to the complexity of the game at the upper echelons of pro play.

Self Milking Cows - How Now, Cow? Dairy scientists use #robotics & cows to decide when to get milked. When ready, they walk into the milking stall, then the robotic arm attaches to begin milking.

The 2019 Major League Baseball (MLB) postseason is here after an exhilarating regular season in which fans saw many exciting new developments. MLB and Amazon Web Services (AWS) teamed up to develop and deliver three new, real-time machine learning (ML) stats to MLB games: Stolen Base Success Probability, Shift Impact, and Pitcher Similarity Match-up Analysis. These features are giving fans a deeper understanding of America's pastime through Statcast AI, MLB's state-of-the-art technology for collecting massive amounts of baseball data and delivering more insights, perspectives, and context to fans in every way they're consuming baseball games. This post looks at the role machine learning plays in providing fans with deeper insights into the game. We also provide code snippets that show the training and deployment process behind these insights on Amazon SageMaker.

If you shop online or occasionally speak to a voice assistant in the morning, you are already embracing the changes this technology has created. Many people are familiar with the advances of autonomous vehicles or facial recognition technology, and some may be curious, or even anxious, about how they will affect safety or privacy. Make no mistake, AI is a transformative technology that is influencing our daily lives and will touch every sector of the global economy. Whether society and government enable or inhibit the AI race, and the extent to which they do so, will be a critical question of the next decade. Regardless of the answer, the technology will forge ahead.

For the last several years, artificial intelligence (AI) has represented the newest, most rapidly expanding frontier of radiology technology. Expo floors at all the major professional society meetings are full of vendors showcasing AI tools they have developed or integrated into their products, billed as efficiency and time-savings aids to help ease the workload of radiologists who are increasingly bogged down by vast amounts of data. Despite the promises and potential, however, widespread clinical implementation of AI in radiology has yet to occur. Early adopters are providing potential pathways for adoption, and vendors and clinicians continue to work together to ensure AI is actually doing what radiologists need it to do. According to numerous key opinion leaders in the fields of radiology and AI, there are a few main obstacles AI currently faces to widespread adoption.

The Roguelike Celebration happened earlier this month at GitHub in San Francisco. Fiction writer Robin Sloan gave a talk, "Writing with the machine: GPT-2 and text generation" at the conference. In this project, Sloan utilized OpenAI's GPT-2 algorithm to write short fantasy stories with help from the general public. To keep things literary, Sloan created a backstory for these stories. In a nutshell, a group of wizards is simulating quests to defeat the dark lord of the land.

A Russian startup has created a new line of androids it hopes to sell to businesses with heavy customer service needs, like airports, banks, and museums. Built by Promobot, a Russian tech company operating out of a Warminister, Pennsylvania, Robo-C is the'world's first humanoid android, which not only simulates the appearance of a person, but also is able to integrate into business processes.' Robo-C was designed to be able to copy human facial expressions and can move its eyes, eyebrows, lips and other facial muscles, via 16 moving parts, and can adopt 600 different facial expressions. The robot can also speak directly to users and comes with an AI containing 100,000 speech modules, according to a report from CNBC. Robo-C is targeted for a number of different commercial uses, including work as a home companion.