Chipmaker Nvidia set a fresh sales record in the second quarter, surpassing Wall Street expectations for its artificial intelligence chips. But shares of the chip giant still dropped 2.3% in after hours trading, in a sign that investors' worries of an AI bubble and the repercussions of Donald Trump's trade wars are not quelled. Nvidia's financial report was the first test of investor appetite since last week's mass AI-stock selloff, when several tech stocks saw shares tumble last week amid growing questions over whether AI-driven companies are being overvalued. On Wednesday, Nvidia reported an adjusted earnings per share of 1.08 on 46.74bn in revenue, surpassing Wall Street's projection of 1.01 in earnings per share on 46.05bn in revenue, according to Fact Set data. But investors had high expectations for the company.

It is the largest collection of video games in the world, worth a whopping $2.1 million (£1.69 million) - but the owner insists he is not done and plans to build on his expensive hobby. Antonio Romero Monteiro, 45, from Texas, USA, has broken the Guinness World record for the largest collection of video games. The avid gamer, whose multi-million-dollar collection began in 1987 when he was just 10-years-old, has collected a whopping 24,268 video games. Mr Monteiro now holds the record for largest collection of Xbox items, largest collection of Sega items, largest collection of Nintendo items, largest collection of PlayStation items and largest collection of video games overall. Antonio Romero Monteiro, 45, (pictured) from Texas, USA, has broken the Guinness World record for the largest collection of video games, worth $2.1 million He owns 24,268 video games, worth $2.1 million, and holds five Guinness World Records: Since buying his first Sega Genesis game more than 30 years ago, his burgeoning collection has expanded to fill a whole room, meticulously organised by system and then alphabetically.

Artificial intelligence and machine learning has the potential to boost many, many areas of the enterprise. As explored in my recent post, it is capable of accelerating and adding intelligence to supply chain management, human resources, sales, marketing and finance. The inevitable impact of AI on IT departments was touched on in a recent survey of 2,280 business leaders from MIT Sloan Management Review and SAS, which finds that in these early days of AI, IT professionals will be feeling the greatest impact -- both from a career and an operational point of view.. CIOs, chief data officers, and chief analytics officers will be on the front lines of AI implementations, the study finds. IT road maps, software development, deployment processes, and data environments are likely to be transformed in the near future. Most IT managers report that they are still developing foundational capabilities for AI -- cloud or data center infrastructure, cybersecurity, data management, development processes and workflow.

Robustness recently becomes one of the major concerns among machine learning community, since learning algorithms are usually vulnerable to outliers or corruptions. Motivated by such a trend and needs, we pursue robustness in semi-definite programming (SDP) in this paper. Specifically, this is done by replacing the commonly used squared loss with the more robust $\ell_1$-loss in the low-rank SDP. However, the resulting objective becomes neither convex nor smooth. As no existing algorithms can be applied, we design an efficient algorithm, based on majorization-minimization, to optimize the objective. The proposed algorithm not only has cheap iterations and low space complexity but also theoretically converges to some critical points. Finally, empirical study shows that the new objective armed with proposed algorithm outperforms state-of-the-art in terms of both speed and accuracy.

We discuss two intrinsic weaknesses of the spectral graph partitioning method, both of which have practical consequences. The first is that spectral embeddings tend to hide the best cuts from the commonly used hyperplane rounding method. Rather than cleaning up the resulting suboptimal cuts with local search, we recommend the adoption of flow-based rounding. The second weakness is that for many "power law" graphs, the spectral method produces cuts that are highly unbalanced, thus decreasing the usefulness of the method for visualization (see figure 4(b)) or as a basis for divide-and-conquer algorithms. These balance problems, which occur even though the spectral method's quotient-style objective function does encourage balance, can be fixed with a stricter balance constraint that turns the spectral mathematical program into an SDP that can be solved for million-node graphs by a method of Burer and Monteiro.

We discuss two intrinsic weaknesses of the spectral graph partitioning method, both of which have practical consequences. The first is that spectral embeddings tend to hide the best cuts from the commonly used hyperplane rounding method. Rather than cleaning up the resulting suboptimal cuts with local search, we recommend the adoption of flow-based rounding. The second weakness is that for many "power law" graphs, the spectral method produces cuts that are highly unbalanced, thus decreasing the usefulness of the method for visualization (see figure 4(b)) or as a basis for divide-and-conquer algorithms. These balance problems, which occur even though the spectral method's quotient-style objective function does encourage balance, can be fixed with a stricter balance constraint that turns the spectral mathematical program into an SDP that can be solved for million-node graphs by a method of Burer and Monteiro.

We discuss two intrinsic weaknesses of the spectral graph partitioning method, both of which have practical consequences. The first is that spectral embeddings tend to hide the best cuts from the commonly used hyperplane rounding method. Rather than cleaning up the resulting suboptimal cutswith local search, we recommend the adoption of flow-based rounding. The second weakness is that for many "power law" graphs, the spectral method produces cuts that are highly unbalanced, thus decreasing theusefulness of the method for visualization (see figure 4(b)) or as a basis for divide-and-conquer algorithms. These balance problems, which occur even though the spectral method's quotient-style objective function does encourage balance, can be fixed with a stricter balance constraint thatturns the spectral mathematical program into an SDP that can be solved for million-node graphs by a method of Burer and Monteiro.