In this paper we present a general framework for estimating regression models subject to a user-defined level of fairness. We enforce fairness as a model selection step in which we choose the value of a ridge penalty to control the effect of sensitive attributes. We then estimate the parameters of the model conditional on the chosen penalty value. Our proposal is mathematically simple, with a solution that is partly in closed form, and produces estimates of the regression coefficients that are intuitive to interpret as a function of the level of fairness. Furthermore, it is easily extended to generalised linear models, kernelised regression models and other penalties; and it can accommodate multiple definitions of fairness. We compare our approach with the regression model from Komiyama et al. (2018), which implements a provably-optimal linear regression model; and with the fair models from Zafar et al. (2019). We evaluate these approaches empirically on six different data sets, and we find that our proposal provides better goodness of fit and better predictive accuracy for the same level of fairness. In addition, we highlight a source of bias in the original experimental evaluation in Komiyama et al. (2018).

The central object of this PhD thesis is known under different names in the fields of computer science and statistical mechanics. In computer science, it is called the Maximum Cut problem, one of the famous twenty-one Karp's original NP-hard problems, while the same object from Physics is called the Ising Spin Glass model. This model of a rich structure often appears as a reduction or reformulation of real-world problems from computer science, physics and engineering. However, solving this model exactly (finding the maximal cut or the ground state) is likely to stay an intractable problem (unless $\textit{P} = \textit{NP}$) and requires the development of ad-hoc heuristics for every particular family of instances. One of the bright and beautiful connections between discrete and continuous optimization is a Semidefinite Programming-based rounding scheme for Maximum Cut. This procedure allows us to find a provably near-optimal solution; moreover, this method is conjectured to be the best possible in polynomial time. In the first two chapters of this thesis, we investigate local non-convex heuristics intended to improve the rounding scheme. In the last chapter of this thesis, we make one step further and aim to control the solution of the problem we wanted to solve in previous chapters. We formulate a bi-level optimization problem over the Ising model where we want to tweak the interactions as little as possible so that the ground state of the resulting Ising model satisfies the desired criteria. This kind of problem arises in pandemic modeling. We show that when the interactions are non-negative, our bi-level optimization is solvable in polynomial time using convex programming.

Subjective texts have been especially studied by several works as they can induce certain behaviours in their users. Most work focuses on user-generated texts in social networks, but some other texts also comprise opinions on certain topics and could influence judgement criteria during political decisions. In this work, we address the task of Targeted Sentiment Analysis for the domain of news headlines, published by the main outlets during the 2019 Argentinean Presidential Elections. For this purpose, we present a polarity dataset of 1,976 headlines mentioning candidates in the 2019 elections at the target level. Preliminary experiments with state-of-the-art classification algorithms based on pre-trained linguistic models suggest that target information is helpful for this task. We make our data and pre-trained models publicly available.

Learning-to-rank, a machine learning technique widely used in information retrieval, has recently been applied to the problem of ligand-based virtual screening, to accelerate the early stages of new drug development. Ranking prediction models learn based on ordinal relationships, making them suitable for integrating assay data from various environments. Existing studies of rank prediction in compound screening have generally used a learning-to-rank method called RankSVM. However, they have not been compared with or validated against the gradient boosting decision tree (GBDT)-based learning-to-rank methods that have gained popularity recently. Furthermore, although the ranking metric called Normalized Discounted Cumulative Gain (NDCG) is widely used in information retrieval, it only determines whether the predictions are better than those of other models. In other words, NDCG is incapable of recognizing when a prediction model produces worse than random results. Nevertheless, NDCG is still used in the performance evaluation of compound screening using learning-to-rank. This study used the GBDT model with ranking loss functions, called lambdarank and lambdaloss, for ligand-based virtual screening; results were compared with existing RankSVM methods and GBDT models using regression. We also proposed a new ranking metric, Normalized Enrichment Discounted Cumulative Gain (NEDCG), which aims to properly evaluate the goodness of ranking predictions. Results showed that the GBDT model with learning-to-rank outperformed existing regression methods using GBDT and RankSVM on diverse datasets. Moreover, NEDCG showed that predictions by regression were comparable to random predictions in multi-assay, multi-family datasets, demonstrating its usefulness for a more direct assessment of compound screening performance.

Developing an understanding of the public discourse on COVID-19 vaccination on social media is important not only for addressing the current COVID-19 pandemic, but also for future pathogen outbreaks. We examine a Twitter dataset containing 75 million English tweets discussing COVID-19 vaccination from March 2020 to March 2021. We train a stance detection algorithm using natural language processing (NLP) techniques to classify tweets as `anti-vax' or `pro-vax', and examine the main topics of discourse using topic modelling techniques. While pro-vax tweets (37 million) far outnumbered anti-vax tweets (10 million), a majority of tweets from both stances (63% anti-vax and 53% pro-vax tweets) came from dual-stance users who posted both pro- and anti-vax tweets during the observation period. Pro-vax tweets focused mostly on vaccine development, while anti-vax tweets covered a wide range of topics, some of which included genuine concerns, though there was a large dose of falsehoods. A number of topics were common to both stances, though pro- and anti-vax tweets discussed them from opposite viewpoints. Memes and jokes were amongst the most retweeted messages. Whereas concerns about polarisation and online prevalence of anti-vax discourse are unfounded, targeted countering of falsehoods is important.

With the development of social media, social communication has changed. While this facilitates people's communication and access to information, it also provides an ideal platform for spreading rumors. In normal or critical situations, rumors will affect people's judgment and even endanger social security. However, natural language is high-dimensional and sparse, and the same rumor may be expressed in hundreds of ways on social media. As such, the robustness and generalization of the current rumor detection model are put into question. We proposed a novel \textbf{h}ierarchical \textbf{a}dversarial \textbf{t}raining method for \textbf{r}umor \textbf{d}etection (HAT4RD) on social media. Specifically, HAT4RD is based on gradient ascent by adding adversarial perturbations to the embedding layers of post-level and event-level modules to deceive the detector. At the same time, the detector uses stochastic gradient descent to minimize the adversarial risk to learn a more robust model. In this way, the post-level and event-level sample spaces are enhanced, and we have verified the robustness of our model under a variety of adversarial attacks. Moreover, visual experiments indicate that the proposed model drifts into an area with a flat loss landscape, leading to better generalization. We evaluate our proposed method on three public rumors datasets from two commonly used social platforms (Twitter and Weibo). Experiment results demonstrate that our model achieves better results than state-of-the-art methods.

The Capture, the BBC's Black Mirror-esque tech thriller-come-police drama, is back for a second season, and the stakes have never been higher. Like the season one premiere, we open with a crime that'll set the tone for the episode to come -- this time, the assassination of a Chinese computer scientist, Edison Yao (Joshua Jo), with apparent links to an AI surveillance firm trying to sell its software to the British government. But it gets weirder: there's CCTV all over his flat and apartment block, and the images pick up nothing but elevators moving and sliding doors parting, as if used by a ghost. That's just where the fun begins. We're then introduced to security minister Isaac Turner MP (Paapa Essiedu), a rare likeable politician placed at the head of a parliamentary committee set up to assess whether the UK Government should use Chinese tech.

Thanks to the recent efforts of such figures as Democratic presidential candidate Andrew Yang and British Shadow Chancellor John McDonnell, the issue of Universal Basic Income (UBI) has been back at the forefront of the public discussion on economic issues, along with the various arguments and justifications for introducing such a policy. While many of these justifications have become quite familiar over the years of waxing and waning interest in UBI, it is interesting to note the recent surge of interest in one particular argument, which sounds more like something from a science fiction novel than an economics textbook. This argument runs roughly as follows: In the not too distant future, rapidly advancing technology will allow robots and artificial intelligence (AI) to perform many of the jobs now being done by humans and to do so more cheaply and efficiently than humans ever could. This will result in robots/AI replacing humans in almost all jobs, making the vast majority of people permanently unemployed, and without Universal Basic Income, how will they (the people) be able to keep food on their tables? Of course, the idea that advances in labor-saving technology will lead to catastrophic unemployment and declining living standards is hardly new, arguably dating back to ancient Greece or earlier, and economists (not to mention the facts of history) have been refuting the idea for nearly as long as economics has existed as a self-conscious science.

The Taliban's acting defence minister has said Pakistan allowed American drones to use its airspace to access Afghanistan, a charge Pakistan has recently denied following a US air strike in Kabul. Acting Minister of Defence Mullah Mohammad Yaqoob told reporters at a news conference in Kabul on Sunday that American drones have been entering Afghanistan via Pakistan. "According to our information the drones are entering through Pakistan to Afghanistan, they use Pakistan's airspace, we ask Pakistan, don't use your airspace against us," he said. Pakistan's foreign ministry did not immediately respond to a request for comment. Pakistani authorities have denied involvement in or advanced knowledge of a drone strike the United States said it carried out in Kabul in July that killed al-Qaeda leader Ayman al-Zawahiri.

Compounding the danger is the lack of any AI regulation. Instead, unaccountable technology conglomerates, such as Google and Meta, have assumed the roles of judge and jury in all things AI. They are silencing dissenting voices, including their own engineers who warn of the dangers. The world's failure to rein in the demon of AI--or rather, the crude technologies masquerading as such--should serve to be a profound warning. There is an even more powerful emerging technology with the potential to wreak havoc, especially if it is combined with AI: quantum computing.