These questions originally appeared on Quora - the knowledge sharing network where compelling questions are answered by people with unique insights. Q: How should one start a career in machine learning? A: There is not just one way. You can start at any age. Some math background (in linear algebra, statistics, and calculus) is recommended, so take classes on these topics, if possible.

There has been much discussion in the last few days about the announcement by the World Economic Forum that it predicts it will take 117 more years until we achieve gender parity in the workplace. It seems crazy that in today's workforce, which is driving developments like self-driving cars, gaming-genius AI, and making hoverboards a reality, we still don't have gender equality. Research published recently by EY makes a compelling case for businesses to do more in terms of tackling existing inequalities: data shows that more diverse company boards command higher share prices and improved financial performance; balanced leadership increases a company's productivity and nationally a country's GDP can be lifted by reducing the gender gap. Another piece of research that looked at start-ups receiving Series A funding in the Bay Area in 2015, showed that only 8% of firms were led by women - that's 16 out of 204 start-ups. And this figure was down by 30% from the previous year.

Our world as we know it is running on artificial intelligence. We have cars that park themselves, and air traffic control is almost fully automated. Virtually every field has benefited from advances in artificial intelligence, from the military to medicine to manufacturing. However, almost none of the recent advancements in artificial intelligence have advanced the education industry. Why is education lagging behind?

A UF club is teaching students about creating systems with artificial intelligence. Nicholas Kroeger, a UF computer science sophomore, and John Henning, a UF computer science and mathematics junior, recently established an artificial intelligence club after realizing UF doesn't offer courses on AI for undergraduate students. The club, which will meet Wednesdays at 6:30 p.m., teaches students how to build computers and software capable of intelligent behavior. "There's no undergraduate classes that I saw that I could take for AI and there's no AI undergraduate major, which I think is really important," Kroeger, 19, said. "I'd like to create those later on in life if I choose to pursue a Ph.D." Bernard Marger, a UF computer engineering senior, said UF used to have an undergraduate artificial intelligence curriculum, but it discontinued when the professor retired.

Editor's Note: Don't miss our new free on-demand training course about how to create data pipeline applications using Apache Spark – learn more here. Decision trees are widely used for the machine learning tasks of classification and regression. In this blog post, I'll help you get started using Apache Spark's MLlib machine learning decision trees for classification. In general, machine learning may be broken down into two classes of algorithms: supervised and unsupervised. Supervised algorithms use labeled data in which both the input and output are provided to the algorithm.

A new Federal Government report has outlined the future of Australia's workforce and revealed growing demand for professionals in STEM industries. The report by CSIRO and the Australian Computer Society, titled Tomorrow's Digitally Enabled Workforce, identified six megatrends. The trends include continued advances in automation and artificial intelligence; jobs will be more flexible and agile due to digital technology; a requirement for entrepreneurial skills; and an increase in skills and education requirements for many professions. While the report found that 44% of Australian jobs would be impacted by these changes, Andrew Johnson, CEO of the Australian Computer Society and one of the report's authors, said there are numerous opportunities for engineers. 'The intent of this report is to look at a 15- to 20-year timeframe.

It was nearly 30 years ago that I first got infatuated with Artificial Intelligence (AI) and I ended up focusing both my undergraduate and graduate engineering research on applications of Artificial Neural Networks (ANNs). My first two jobs after graduate school stayed in the same groove; over 6 years I developed AI and machine learning techniques to address real world problems that ranged from recognizing human speech and natural language, to converting handwriting to searchable digitized text, and to streamlining maintenance procedures in nuclear reactor cores. So it is with a mix of amazement and amusement that I am soaking up the resurgence of AI and machine learning as the buzzword-du-jour: "Deep Learning". Deep Learning is very visible in the high hopes we hold for driverless cars and in the triumph of machines over chess champions. It is less conspicuously and more frequently used in the form of Apple's Siri, Amazon's Echo, playlists generated on Spotify, that auto-tag feature on Facebook Photos, the voice assistant that answers the phone when you call your bank, or when your fingerprint is recognized by a machine.

The World Science Festival Brisbane's'Shared Space Bots' performance demonstrations at Queensland University of Technology used pint-sized, futuristic floor robots to reveal research into the technologies that will allow humans to communicate with driverless cars, and allowed audience members a chance to'test drive' the systems safely. Shared Space Bots were demonstrated by internationally acclaimed technologist Christopher Lindinger from Austrian R&D company Ars Electronica Futurelab. They have been developed as part of the ongoing research cooperation between Mercedes-Benz and Ars Electronica Futurelab on the topic of future mobility.

The web is full of good explanations of machine learning algorithms. And every second applicant for a data science position has finished the Coursera course on machine learning. Theory will not help you choose good values for the 16 parameters a standard implementation of a random forest takes. The default values are good to get started, but which parameters should you modify depending on your data? Choosing the right features, algorithms and parameters is an art.

Stochastic gradient descent is the method of choice for large-scale machine learning problems, by virtue of its light complexity per iteration. However, it lags behind its non-stochastic counterparts with respect to the convergence rate, due to high variance introduced by the stochastic updates. The popular Stochastic Variance-Reduced Gradient (SVRG) method mitigates this shortcoming, introducing a new update rule which requires infrequent passes over the entire input dataset to compute the full-gradient. In this work, we propose CheapSVRG, a stochastic variance-reduction optimization scheme. Our algorithm is similar to SVRG but instead of the full gradient, it uses a surrogate which can be efficiently computed on a small subset of the input data. It achieves a linear convergence rate ---up to some error level, depending on the nature of the optimization problem---and features a trade-off between the computational complexity and the convergence rate. Empirical evaluation shows that CheapSVRG performs at least competitively compared to the state of the art.