No matter how many books you read, tutorials you finish or problems you solve, there will always be a data set you might come across where you get clueless. Specially, when you are in your early days of Machine Learning. In this blog post, you'll learn some essential tips on building machine learning models which most people learn with experience. These tips were shared by Marios Michailidis (a.k.a Kazanova), Kaggle Grandmaster, Current Rank #3 in a webinar happened on 5th March 2016. The key to succeeding in competitions is perseverance. Marios said, 'I won my first competition (Acquired valued shoppers challenge) and entered kaggle's top 20 after a year of continued participation on 4 GB RAM laptop (i3)'. Were you planning to give up? While reading Q & As, if you have any questions, please feel free to drop them in comments!

SAP Labs in India is the second largest R&D centre for the company after its centre in Walldorf in Germany and among the 3 hubs in SAP Labs network of 19 Labs across 16 countries. Dilipkumar Khandelwal, Managing Director for SAP Labs in India, has a dual role as he is also the Executive Vice President and Global Head of Enterprise Cloud Services for SAP. Edited excerpts: What is the SAP Lab's focus here? Do you focus more on the local needs to serve SAP India or do you essentially serve global needs? SAP Labs India was founded in 1998.

His high school programming club was arguing about whether artificial intelligence could ever accomplish tasks better than humans. Barrat thought the answer was obvious. A few of his peers, however, weren't so easily convinced and asked for proof by the club's next meeting. "All of the sudden I had a week to make a neural network that could rap," Barrat said. Barrat's story is possible because Silicon Valley has decided AI is becoming indispensable, and big companies need to cultivate more talent to fill the growing demand--Google, Facebook, Microsoft, IBM and other giants like GE are shelling out multi-million dollar salaries for AI programmers. To upend the perceived shortage of talent, tech companies have begun to evangelize for open-source AI code, or software that's free to use, modify, and improve upon.

Autonomous vehicles are already a common sight on the streets of Silicon Valley, an international hub for self-driving technology. But this month, California set the stage for the next phase of innovation that could dramatically alter transportation and mobility across the globe. The state has proposed regulations to allow fully autonomous vehicles to drive on public roads – meaning empty cars with no steering wheels and no backup driver inside. The new rules are a game-changer for the nascent industry, opening the doors to a host of complex questions about legality, ethics and safety. The regulations, which could go into effect this year, pave the road for a deployment that could revolutionize modern society.

Machine learning is typified by algorithms that are capable of deriving patterns and'intelligence' from vast quantities of training data. As company's such as DeepMind are beginning to show us however, the real improvements come when the machines are capable of learning independently of data fed it by us. A recent paper from researchers at the University of Toronto highlights some of the progress being made. The team developed an algorithm that attempts to learn from human instructions. In other words, it doesn't require a database of past examples, as is the norm today.

Artificial intelligence is shaping up as the next industrial revolution, poised to rapidly reinvent business, the global economy and how people work and interact with each other. Andrew Ng, chief scientist at Chinese internet giant Baidu Inc. and co-founder of education startup Coursera, and Neil Jacobstein, chair of the artificial intelligence and robotics department at Silicon Valley think tank Singularity University, sat down with The Wall Street Journal's Scott Austin to discuss AI's opportunities and challenges. What is Baidu focused on? NG: For large enterprises like Baidu, AI creates two big pockets of opportunities. One is our core business.

When applied to the S&P 500 constituents from 1992 to 2015, their stock selections generated annual returns in the double digits -- whereas the highest profits were made at times of financial turmoil. In March 2016, South Korean Lee Sedol, one of the best Go players in the world, lost to the AlphaGo computer program. It was a milestone in the history of artificial intelligence because up to that point the Asian board game had been considered too complex for computers. Behind successes such as this are programs that are modelled on biological systems and are constructed in a form similar to neural networks so that they can independently extract relationships from millions of data points. 'Artificial neural networks are primarily applied to problems, where solutions cannot be formulated with explicit rules,' explains Dr. Christopher Krauss of the Chair for Statistics and Econometrics at FAU. 'Image and speech recognition are typical fields of application, such as Apple's Siri.

This playlist contains the lecture videos for the Deep Natural Language Processing course offered in Hilary Term 2017 at the University of Oxford. This is an advanced course on natural language processing. Automatically processing natural language inputs and producing language outputs is a key component of Artificial General Intelligence. The ambiguities and noise inherent in human communication render traditional symbolic AI techniques ineffective for representing and analysing language data. Recently statistical techniques based on neural networks have achieved a number of remarkable successes in natural language processing leading to a great deal of commercial and academic interest in the field This is an applied course focusing on recent advances in analysing and generating speech and text using recurrent neural networks.

A common problem in large-scale data analysis is to approximate a matrix using a combination of specifically sampled rows and columns, known as CUR decomposition. Unfortunately, in many real-world environments, the ability to sample specific individual rows or columns of the matrix is limited by either system constraints or cost. In this paper, we consider matrix approximation by sampling predefined blocks of columns (or rows) from the matrix. This regime is commonly found when data is distributed across multiple nodes in a compute cluster, where such blocks correspond to columns (or rows) of the matrix stored on the same node, which can be retrieved with much less overhead than retrieving individual columns stored across different nodes. We propose a novel algorithm for sampling useful column blocks and provide guarantees for the quality of the approximation. We demonstrate the practical utility of this algorithm for computing the block CUR decomposition of large matrices in a distributed setting using Apache Spark. Using our proposed block CUR algorithms, we can achieve a significant speed-up compared to a regular CUR decomposition with the same quality of approximation.

Bandit methods for black-box optimisation, such as Bayesian optimisation, are used in a variety of applications including hyper-parameter tuning and experiment design. Recently, \emph{multi-fidelity} methods have garnered considerable attention since function evaluations have become increasingly expensive in such applications. Multi-fidelity methods use cheap approximations to the function of interest to speed up the overall optimisation process. However, most multi-fidelity methods assume only a finite number of approximations. In many practical applications however, a continuous spectrum of approximations might be available. For instance, when tuning an expensive neural network, one might choose to approximate the cross validation performance using less data $N$ and/or few training iterations $T$. Here, the approximations are best viewed as arising out of a continuous two dimensional space $(N,T)$. In this work, we develop a Bayesian optimisation method, BOCA, for this setting. We characterise its theoretical properties and show that it achieves better regret than than strategies which ignore the approximations. BOCA outperforms several other baselines in synthetic and real experiments.