Heterogeneous unsupervised domain adaptation (HUDA) is the most challenging domain adaptation setting where the feature spaces of source and target domains are heterogeneous, and the target domain has only unlabeled data. Existing HUDA methods assume that both positive and negative examples are available in the source domain, which may not be satisfied in some real applications. This paper addresses a new challenging setting called positive and unlabeled heterogeneous unsupervised domain adaptation (PU-HUDA), a HUDA setting where the source domain only has positives. PU-HUDA can also be viewed as an extension of PU learning where the positive and unlabeled examples are sampled from different domains. A naive combination of existing HUDA and PU learning methods is ineffective in PU-HUDA due to the gap in label distribution between the source and target domains. To overcome this issue, we propose a novel method, predictive adversarial domain adaptation (PADA), which can predict likely positive examples from the unlabeled target data and simultaneously align the feature spaces to reduce the distribution divergence between the whole source data and the likely positive target data. PADA achieves this by a unified adversarial training framework for learning a classifier to predict positive examples and a feature transformer to transform the target feature space to that of the source. Specifically, they are both trained to fool a common discriminator that determines whether the likely positive examples are from the target or source domain. We experimentally show that PADA outperforms several baseline methods, such as the naive combination of HUDA and PU learning.

For years, anti-doping authorities have struggled against the ever-evolving campaigns cheaters use to gain an advantage. Cycling's history can be traced back by the developments in banned substances, from amphetamines to blood boosters and transfusions. While those chasing the dopers have been able to evolve their methods in response to the changing tides, the use of performance enhancing drugs continues to blight the cycling world, as proven by the recent Operation Aderlass blood doping scandal. But the World Anti-Doping Agency (WADA) is turning its eye to a new method of detecting who is using clandestine methods to gain an advantage – artificial intelligence. WADA and the Fonds de recherche du Québec (Québec research fund) announced this week that it has handed over funding to three separate projects that will explore the possible uses of AI in the fight against doping.

The World Anti-Doping Agency plans to use artificial intelligence in its fight against doping, Olivier Niggli, the organisation's director general has exclusively revealed to i. There's a lot of promising things," Niggli said. Wada will launch a call for pilot artificial intelligence projects in the coming weeks, as it intensifies attempts to use the technology. The organisation believe that artificial intelligence can be used to identify suspicious athletes, raise red flags and improve how testing is targeted. Under the plans for how to use the technology, if an athlete was flagged by artificial intelligence, it would trigger immediate additional targeted testing. Wada intend to use artificial intelligence to identify patterns in the vast amounts of data that anti-doping bodies already collect.