Data Collection Guide: Fill & Download for Free

A friend and I put together a reading list internally -- here you go:Prerequisites: basic RL (estimated time to complete: 1 day)Sutton and Barto’s book, Section I (The Problem) and Section II (Elementary Solution Methods)Reinforcement Learning: An IntroductionDavid Silver’s course, lectures 1-7Advanced Topics: RLFundamental algorithmsDeep Q learningContinuous Deep Q-Learning with Model-based Acceleration: [1603.00748] Continuous Deep Q-Learning with Model-based AccelerationPolicy gradientTrust Region Policy Optimization: [1502.05477] Trust Region Policy OptimizationVariance reduction: Variance Reduction Techniques for Gradient Estimates in Reinforcement LearningDeterministic policy gradient: http://jmlr.org/proceedings/papers/v32/silver14.pdf [1509.02971] Continuous control with deep reinforcement learningGuided policy search:Learning Complex Neural Network Policies with Trajectory Optimization: http://www.eecs.berkeley.edu/~svlevine/papers/cgps.pdfEnd-to-End training of Deep Visuomotor Policies: http://arxiv.org/pdf/1504.00702v1.pdfLearning to search: Dagger/Aggrevate:https://www.cs.cmu.edu/~sross1/publications/Ross-AIStats11-NoRegret.pdf[1406.5979] Reinforcement and Imitation Learning via Interactive No-Regret LearningImpressive ApplicationsAlphaGo + MCTShttp://www.nature.com/nature/journal/v529/n7587/full/nature16961.htmlA survey of Monte Carlo Tree Search Methods: A Survey of Monte Carlo Tree Search MethodsLearning Hand-Eye Coordination for Robotic Grasping with Deep Learning and Large-Scale Data Collection: [1603.02199] Learning Hand-Eye Coordination for Robotic Grasping with Deep Learning and Large-Scale Data CollectionOn the horizon: Intrinsic motivation, exploration (e.g. http://arxiv.org/pdf/1602.04621)Multiplayer gamesLearning approximate Nash equilibria in imperfect-information games [1603.01121] Deep Reinforcement Learning from Self-Play in Imperfect-Information Games

There are a number of great apps for gather data from a form via a mobile device. Essentially, you'll make a form online, then deploy it to the form tool's mobile app where you can fill it out online or offline. Most will let you gather images and geolocation data, scan barcodes, view documentation, and more, making them far more useful and powerful than your standard online form apps.We've recently rounded up some of the best mobile data collection apps at Zapier. You can read the entire post at Zapier (https://zapier.com/learn/ultimate-guide-to-forms-and-surveys/best-data-collection-apps/) or here's some of the apps to explore:Device Magic for pre-filled data in formsGoFormz for digitizing paper formsNexticy for paper-style forms on iPadQuickTapSurvey for questions and surveysFulcrum for geolocation with custom mapsFormitize for accepting payments in mobile formsiFormBuilder for adding data to your database from anywherePushForms for turning paper forms into mobile appsFormhub for open source data collectionTrackVia for building powerful apps that work anywhereAlso, for more tips on using forms, check out our free eBook "The Ultimate Guide to Forms and Surveys".

The current state of AI is nothing like what you expect from Sci-fi. It's much more like advanced pattern detection programs than some sort of mastermind.The idea of an AI guiding national policy is just science fiction.There may be some targeted use of AI which affect national policy, for instance looking for patterns indicating possible terrorist activity in the vast amount of data collected by the NSA.Like I said, think pattern matching , not super mind.