To do capable and useful things, your robot needs capable and useful sensors, which is just another way of saying that your robot needs you to spend a lot of money on it. This is really too bad, because hardware cost is enormously restrictive for robots, especially ones that are intended to be affordable by people who haven't co-founded a robotics startup or something (I think there are a few people left who have yet to do this). In particular, distance sensors that allow your robot to detect and avoid obstacles tend to be both very useful and very expensive, but if you want one that works reliably outdoors, start saving, because they cost thousands of dollars. At MIT, a group of researchers led by Professor Li-Shiuan Peh designed a phone-based laser rangefinder that costs a total of 49, plus a smartphone that you're not using anymore. Is it the greatest laser rangefinder ever?

In episode six of season two, we talk about how to build software for machine learning (and what the roadblocks are), we take a listener question about how to start exploring a new dataset, plus, we talk with Rob Tibshirani of Stanford University.

Editor's note: This is one in a series of profiles of five finalists for NVIDIA's 2016 Global Impact Award, which provides 150,000 to researchers using NVIDIA technology for groundbreaking work that addresses social, humanitarian and environmental problems. Eradicating worldwide poverty by 2030 is the top goal on the United Nations' sustainable development agenda, published late last year. But a lack of data has frustrated efforts to measure progress toward the goal. Most of those living in extreme poverty are in sub-Saharan Africa and Southern Asia, where accurate poverty data is scarce. A small team at Stanford University is changing that, one satellite image at a time.

Microsoft's annual mega-gathering of developers kicked-off Wednesday in San Francisco, where the company presented its latest tools and technologies. The company claimed that Windows 10 is off to the fastest start in its history with over 270 million active devices. And, as a result of this rapidly growing base, Microsoft is seeing new universal apps from Twitter, Uber, King, Disney, Wargaming, Square Enix, Yahoo and WWE; with new apps on the way from Bank of America, Starbucks, Facebook, Messenger and Instagram. "As an industry, we are on the cusp of a new frontier that pairs the power of natural human language with advanced machine intelligence," Microsoft CEO Satya Nadella said in his keynote address to thousands of developers at Build 2016. The giant announced new additions to the Cortana Intelligence Suite, formerly known as the Cortana Analytics Suite, which is powered by cutting-edge research into big data, machine learning, perception, analytics and intelligent bots.

According to Okta CEO Todd McKinnon, there's a lot of hype around the potential of machine learning, but companies aren't actually taking advantage of it. It's similar to how people discussed big data a few years ago. In his view, tech companies need to create and sell intelligent services that let other businesses use machine learning to perform key tasks. Microsoft's data chief, Joseph Sirosh, has said he expects to see a marketplace of intelligent algorithms and applications that companies can buy.

A computer program has defeated a master of the ancient Chinese game of Go, achieving one of the loftiest of the Grand Challenges of AI at least a decade earlier than anyone had thought possible. The programmers, at Google's Deep Mind laboratory, in London, write in today's issue of Nature that their program AlphaGo defeated Fan Hui, the European Go champion, 5 games to nil, in a match held last October in the company's offices. Earlier, the program had won 494 out of 495 games against the best rival Go programs. AlphaGo's creators now hope to seal their victory at a 5-game match against Lee Se-dol, the best Go player in the world. That match, for a 1 million prize fund, is scheduled to take place in March in Seoul, South Korea.

With hosts and guests around the globe, Airbnb aspires to provide a frictionless payments experience where our guests can pay in their local currency via a familiar payment method, and our hosts can receive money via convenient means in their preferred currency. For example, in Brazil the currency is Brazilian Real, and people are familiar with Boleto as a payment method. These are quite different than what we use in the US, and imagine this problem spread across the 190 countries Airbnb serves. In order to achieve this, our Payments team has built a world-class payments platform that is secure and easy to use. The team's responsibilities include support of guest payments and host payouts, new payment experiences like gift cards, and assisting in financial reconciliation, to name a few.

When it comes to machine learning, the future is already here, but it's not yet evenly distributed. Taking advantage of breakthroughs in the field can require a lot of work, which is tough for small companies and those without a whole team to build custom applications and algorithms. It's similar to how people discussed big data a few years ago. "We think about this a lot, and the most interesting thing about machine learning that I've noticed over the last year is that it's kind of like what big data was three years ago," he said. "Everyone talks about it, but nobody really has it."

If you have been using GBM as a'black box' till now, may be it's time for you to open it and see, how it actually works! This article is inspired by Owen Zhang's (Chief Product Officer at DataRobot and Kaggle Rank 3) approach shared at NYC Data Science Academy. He delivered a 2 hours talk and I intend to condense it and present the most precious nuggets here. Boosting algorithms play a crucial role in dealing with bias variance trade-off. Unlike bagging algorithms, which only controls for high variance in a model, boosting controls both the aspects (bias & variance), and is considered to be more effective.

We have several machine learning algorithms at our disposal for model building. Doing data based prediction is now easier like never before. Whether it is a regression or classification problem, one can effortlessly achieve a reasonably high accuracy using a suitable algorithm. But, this is not the case everytime. Classification problems can sometimes get a bit tricky. ML algorithms tend to tremble when faced with imbalanced classification data sets. Moreover, they result in biased predictions and misleading accuracies. But, why does it happen? What factors deteriorate their performance?