Long-range drone and missile strikes on Russian soil have shifted the balance of the conflict--will they be enough to end it? In recent weeks, Donald Trump--whose attitude toward Ukraine and its President, Volodymyr Zelensky, has been notoriously fickle, when not outright hostile --has appeared to warm to Ukraine's chances in the war with Russia. On the sidelines of the summit in Ankara, he met with Zelensky and later spoke of him in flattering terms, declaring "We've actually developed a good relationship. Trump has a history of favoring those he perceives as winners, and his change in attitude regarding Ukraine seems driven by the country's newfound ability to strike targets deep inside Russia--"an escalation that can help lead to an end," Trump said. In Ankara, Zelensky, the inveterate entertainer, riffed on a question from Trump, who asked if Zelensky would consider travelling to Moscow for direct talks with Russian President Vladimir Putin. That would be dangerous, Zelensky ...

Is the war entering a new phase? Russian drones and missiles killed four people in Ukraine overnight, while Ukrainian attacks on Russia and Russian-occupied areas of Ukraine killed five. Three people were killed in Russian attacks on Ukraine's central Dnipropetrovsk region overnight, including two at an "industrial enterprise" in the city of Kryvyi Rih, regional officials said on Sunday. Russia has escalated attacks in recent weeks, taking advantage of Ukraine's critical shortage of munitions for its Patriot air defence system, which has rendered it largely unable to intercept ballistic missiles flying at several times the speed of sound. NATO countries pledged at their summit in Ankara last week to provide more Patriot munitions to Ukraine, and President Donald Trump said he was willing to give Kyiv a license to manufacture the US missiles domestically.

Donald Trump's NATO summit remarks in Ankara reveal an emerging grand strategy spanning Iran, Ukraine, China, Greenland and defense spending priorities.

NATO Secretary-General Mark Rutte delivers the keynote speech at the NATO Summit Defense Industry Forum, on the sidelines of the NATO leaders' Summit, in Ankara on Tuesday. NATO allies have agreed to at least $50 billion in defense industry deals, according to an alliance official, to show to U.S. President Donald Trump that Europe is heeding his spending demands. Secretary-General Mark Rutte revealed some of the contracts on Tuesday during a defense industry forum in Ankara, where the military alliance's leaders are meeting for their annual summit this week. Those included $12 billion in deals to buy next-generation drones, surveillance planes and military aircraft. Notably, some of the contracts show Europe moving to locally source some equipment it previously bought from the United States.

Ukraine's president plans to use the Nato meeting in Turkey to urge Kyiv's allies to deliver the air defence systems it urgently needs to protect it from escalating Russian attacks. Volodymyr Zelensky's call for help rings with extra intensity after Russian missiles rained down on the Ukrainian capital twice in less than a week, crashing into blocks of flats and killing more than 50 civilians. The summit in Ankara will also be a chance for Zelensky to hold a crucial meeting with Donald Trump and press home his case that Russia's brutal attacks are a show of weakness, not strength, and that Vladimir Putin should be pressured into talks towards a dignified peace. The latest strikes on Ukraine come as it has been stepping up its own long-range drone attacks against Russia, hitting oil refineries and military targets there and causing significant fuel shortages and power cuts. To play this video you need to enable JavaScript in your browser. Russian social media accounts are full of videos of people queuing for hours to buy petrol and fighting over what little they're allowed.

What are Russia's gains from the Iran war? 'We are not losers; we are winners' Russia strikes Ukraine's Odesa port, kills railway worker in Zaporizhia Russian drones have attacked Ukraine's main Black Sea port in the southern city of Odesa and a railway in the region of Zaporizhia, killing a train driver, according to Ukraine's Deputy Prime Minister Oleksii Kuleba. The overnight attacks damaged the infrastructure of the Odesa port, including berths, warehouses, railway infrastructure and port operators' facilities, Kuleba said in a statement on X on Wednesday. Kuleba said this is "another proof of terrorism, Russia is at war against peaceful people, against those who were simply doing their job and keeping the country moving". Russia also launched several drones and missiles on a flight path near the disused Chornobyl nuclear plant, elevating the risk of a significant accident, according to Ukraine's top state prosecutor. This comes as Ukraine prepares to mark the 40th anniversary of the 1986 Chornobyl disaster on Sunday.

We study downlink beam and rate adaptation in a multi-user mmWave MISO system where multiple base stations (BSs), each using analog beamforming from finite codebooks, serve multiple single-antenna user equipments (UEs) with a unique beam per UE and discrete data transmission rates. BSs learn about transmission success based on ACK/NACK feedback. To encode service goals, we introduce a satisficing throughput threshold $τ_r$ and cast joint beam and rate adaptation as a combinatorial semi-bandit over beam-rate tuples. Within this framework, we propose SAT-CTS, a lightweight, threshold-aware policy that blends conservative confidence estimates with posterior sampling, steering learning toward meeting $τ_r$ rather than merely maximizing. Our main theoretical contribution provides the first finite-time regret bounds for combinatorial semi-bandits with satisficing objective: when $τ_r$ is realizable, we upper bound the cumulative satisficing regret to the target with a time-independent constant, and when $τ_r$ is non-realizable, we show that SAT-CTS incurs only a finite expected transient outside committed CTS rounds, after which its regret is governed by the sum of the regret contributions of restarted CTS rounds, yielding an $O((\log T)^2)$ standard regret bound. On the practical side, we evaluate the performance via cumulative satisficing regret to $τ_r$ alongside standard regret and fairness. Experiments with time-varying sparse multipath channels show that SAT-CTS consistently reduces satisficing regret and maintains competitive standard regret, while achieving favorable average throughput and fairness across users, indicating that feedback-efficient learning can equitably allocate beams and rates to meet QoS targets without channel state knowledge.

Preprocessing leakage arises when scaling, imputation, or other data-dependent transformations are estimated before resampling, inflating apparent performance while remaining hard to detect. We present fastml, an R package that provides a single-call interface for leakage-aware machine learning through guarded resampling, where preprocessing is re-estimated inside each resample and applied to the corresponding assessment data. The package supports grouped and time-ordered resampling, blocks high-risk configurations, audits recipes for external dependencies, and includes sandboxed execution and integrated model explanation. We evaluate fastml with a Monte Carlo simulation contrasting global and fold-local normalization, a usability comparison with tidymodels under matched specifications, and survival benchmarks across datasets of different sizes. The simulation demonstrates that global preprocessing substantially inflates apparent performance relative to guarded resampling. fastml matched held-out performance obtained with tidymodels while reducing workflow orchestration, and it supported consistent benchmarking of multiple survival model classes through a unified interface.

Our contributions are: (1) We develop a generalized framework to optimize non-differentiable ranking-based functions byextending theerror-drivenoptimization ofAPLoss.(2)Weprovethat