Machine learning is a data science technique that allows computers to use existing data to forecast future behaviors, outcomes, and trends. Using machine learning, computers learn without being explicitly programmed. Azure Machine Learning is a cloud predictive analytics service that makes it possible to quickly create and deploy predictive models as analytics solutions. Join Ryan Berry and Matthew Puckett as they welcome Microsoft Data Scientist extraordinaire Laura Edell as they provide a brief history of Machine Learning and how you can apply it in your daily operations. How can we say what is good vs. what is bad? 0:37:52 – Let's talk about decision trees – how valuable are these to setting up you data model?

Artificial intelligence works better than human lawyers in accurately spotting risks in everyday business contracts. That's the conclusion of a landmark study pitting 20 experienced US-trained corporate lawyers against Tel Aviv based LawGeex, which makes artificial intelligence (AI) software for contract review and approval, in spotting issues in everyday contracts. Both the lawyers and the LawGeex AI analyzed five previously unseen non-disclosure agreement contracts, containing 153 paragraphs of technical legal language under controlled conditions. This is the first time that AI software has been tested with a typical task undertaken by lawyers on a daily basis. The result: LawGeex software achieved a 94 percent accuracy rate at identifying risks in the NDAs.

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society. KASPAR (Kinesics and Synchronization in Personal Assistant Robotics) is a robot originally conceived as part of a research project begun in the late 1990s by artificial intelligence researcher Kerstin Dautenhahn and her collaborators at the University of Reading in England. Initially, the objective was to develop "robotic therapy games" to facilitate communication with autistic children and to help them interact with others. In 2005, now at the University of Hertfordshire, the KASPAR Project was formally launched with the aim of developing a "social" robot having two missions: first, and mainly, to be a "social mediator" responsible for facilitating communication between autistic children and the people with whom they are in daily contact--other children (autistic or not), therapists, teachers, and parents--and also to serve as a therapeutic and learning tool designed to stimulate social development in these children. The objective was to teach young people with autism a variety of skills that most of us master, more or less fully, without any need of special education: understanding others' emotions and reacting appropriately, expressing our own feelings, playing in a group while letting everyone take turns, and imitating and cooperating with others.

The ability to manipulate and understand data is increasingly critical to discovery and innovation. As a result, we see the emergence of a new field--data science--that focuses on the processes and systems that enable us to extract knowledge or insight from data in various forms and translate it into action. In practice, data science has evolved as an interdisciplinary field that integrates approaches from such data-analysis fields as statistics, data mining, and predictive analytics and incorporates advances in scalable computing and data management. But as a discipline, data science is only in its infancy. The challenge of developing data science in a way that achieves its full potential raises important questions for the research and education community: How can we evolve the field of data science so it supports the increasing role of data in all spheres? How do we train a workforce of professionals who can use data to its best advantage? What should we teach them? What can government agencies do to help maximize the potential of data science to drive discovery and address current and future needs for a workforce with data science expertise?

Abstract-- We propose the use of Bayesian networks, which provide both a mean value and an uncertainty estimate as output, to enhance the safety of learned control policies under circumstances in which a test-time input differs significantly from the training set. Our algorithm combines reinforcement learning and end-to-end imitation learning to simultaneously learn a control policy as well as a threshold over the predictive uncertainty of the learned model, with no hand-tuning required. Corrective action, such as a return of control to the model predictive controller or human expert, is taken when the uncertainty threshold is exceeded. We demonstrate that our method is robust to uncertainty resulting from varying system dynamics as well as from partial state observability. As the deployment of deep neural networks as controllers for physical robotic systems becomes more prevalent, the issue of safety within artificial intelligence becomes an increasingly important concern. Recently the use of end-to-end imitation learning to develop neural network control policies has surged in popularity, due in large part to the ease with which deep models can learn complex dynamics and infer global state from local data while bypassing the need for significant parameter tuning. In contrast, traditional approaches to vision-based control rely on methods such image segmentation and object detection, classification, labeling, and filtering; often, these methods require significant engineering and tuning.

Job openings often go unfulfilled despite a surfeit of unemployed or underemployed workers. One of the main reasons for this is a mismatch between the skills required by employers and the skills that workers possess. This mismatch, also known as the skills gap, can pose socioeconomic challenges for an economy. A first step in alleviating the skills gap is to accurately detect skills in human capital data such as resumes and job ads. Comprehensive and accurate detection of skills facilitates analysis of labor market dynamics. It also helps bridge the divide between supply and demand of labor by facilitating reskilling and workforce training programs. In this paper, we describe SKILL, a Named Entity Normalization (NEN) system for occupational skills. SKILL is composed of 1) A skills tagger which uses properties of semantic word vectors to recognize and normalize relevant skills, and 2) A skill entity sense disambiguation component which infers the correct meaning of an identified skill. We discuss the technical design and the synergy between data science and engineering that was required to transform the system from a research prototype to a production service that serves customers from across the organization. We also discuss establishing customer feedback loops, and it led to improvements to the system over time. SKILL is currently used by various internal teams at CareerBuilder for big data workforce analytics, semantic search, job matching, and recommendations.

Recently, AAAI coordinated and The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19) cosigned a statement with CRA, and the Thirty-First Conference on Innovative Applications of Artificial expressing concern about the proposed Intelligence (IAAI-19), will be held in Honolulu, Hawaii, USA, January tax bill and its ramifications for graduate 27 - February 1, 2019. The technical conference will continue its student stipends. Other organizational 3.5-day schedule, preceded by the workshop and tutorial programs.

The Eighth International Conference on Computational Creativity (ICCC’17)1 was hosted at the Georgia Institute of Technology in Atlanta, Georgia, USA from June 19th - June 23rd, 2017. The ICCC’17 organising committee consisted of Ashok Goel (General Chair), Kazjon Grace (Workshop Co-chair), Matthew Guzdial (Media Chair), Mikhail Jacob (Local Chair), Anna Jordanous (Program Co-chair), Ruli Manurung (Workshop Co-chair) and Alison Pease (Program Co-chair). This report summarises the main topics addressed.

The idea of temporal abstraction, i.e. learning, planning and representing the world at multiple time scales, has been a constant thread in AI research, spanning sub-fields from classical planning and search to control and reinforcement learning. For example, programming a robot typically involves making decisions over a set of controllers, rather than working at the level of motor torques. While temporal abstraction is a very natural concept, learning such abstractions with no human input has proved quite daunting. In this paper, we present a general architecture, called option-critic, which allows learning temporal abstractions automatically, end-to-end, simply from the agent’s experience. This approach allows continual learning and provides interesting qualitative and quantitative results in several tasks.

Machine learning workflow development is anecdotally regarded to be an iterative process of trial-and-error with humans-in-the-loop. However, we are not aware of quantitative evidence corroborating this popular belief. A quantitative characterization of iteration can serve as a benchmark for machine learning workflow development in practice, and can aid the development of human-in-the-loop machine learning systems. To this end, we conduct a small-scale survey of the applied machine learning literature from five distinct application domains. We collect and distill statistics on the role of iteration within machine learning workflow development, and report preliminary trends and insights from our investigation, as a starting point towards this benchmark. Based on our findings, we finally describe desiderata for effective and versatile human-in-the-loop machine learning systems that can cater to users in diverse domains.