We study low-rank parameterizations of weight matrices with embedded spectral properties in the Deep Learning context. The low-rank property leads to parameter efficiency and permits taking computational shortcuts when computing mappings. Spectral properties are often subject to constraints in optimization problems, leading to better models and stability of optimization. We start by looking at the compact SVD parameterization of weight matrices and identifying redundancy sources in the parameterization. We further apply the Tensor Train (TT) decomposition to the compact SVD components, and propose a non-redundant differentiable parameterization of fixed TT-rank tensor manifolds, termed the Spectral Tensor Train Parameterization (STTP). We demonstrate the effects of neural network compression in the image classification setting and both compression and improved training stability in the generative adversarial training setting.

The Isomap is a well-known nonlinear dimensionality reduction method that highly suffers from computational complexity. Its computational complexity mainly arises from two stages; a) embedding a full graph on the data in the ambient space, and b) a complete eigenvalue decomposition. Although the reduction of the computational complexity of the graphing stage has been investigated, yet the eigenvalue decomposition stage remains a bottleneck in the problem. In this paper, we propose the Low-Rank Isomap algorithm by introducing a projection operator on the embedded graph from the ambient space to a low-rank latent space to facilitate applying the partial eigenvalue decomposition. This approach leads to reducing the complexity of Isomap to a linear order while preserving the structural information during the dimensionality reduction process. The superiority of the Low-Rank Isomap algorithm compared to some state-of-art algorithms is experimentally verified on facial image clustering in terms of speed and accuracy.

There has been a growing interest in model-agnostic methods that can make deep learning models more transparent and explainable to a user. Some researchers recently argued that for a machine to achieve a certain degree of human-level explainability, this machine needs to provide human causally understandable explanations, also known as causability. A specific class of algorithms that have the potential to provide causability are counterfactuals. This paper presents an in-depth systematic review of the diverse existing body of literature on counterfactuals and causability for explainable artificial intelligence. We performed an LDA topic modelling analysis under a PRISMA framework to find the most relevant literature articles. This analysis resulted in a novel taxonomy that considers the grounding theories of the surveyed algorithms, together with their underlying properties and applications in real-world data. This research suggests that current model-agnostic counterfactual algorithms for explainable AI are not grounded on a causal theoretical formalism and, consequently, cannot promote causability to a human decision-maker. Our findings suggest that the explanations derived from major algorithms in the literature provide spurious correlations rather than cause/effects relationships, leading to sub-optimal, erroneous or even biased explanations. This paper also advances the literature with new directions and challenges on promoting causability in model-agnostic approaches for explainable artificial intelligence.

Although growing numbers of organisations are working with artificial intelligence (AI) software in some shape or form, very few are generating significant financial benefits when rolling it out in a serious way, according to new research. A study conducted by the MIT Sloan Management Review and management consulting firm the Boston Consulting Group revealed that as many as 57% of the 3,000 managers, executives and academics questioned were currently either piloting or deploying the technology. A further 59% had devised an AI strategy and 70% believed they understood how the software could generate business value. Despite this situation, the report, Expanding AI's impact with organizational learning, indicated that just one in 10 organisations were deriving significant financial value from the technology. When exploring the reasons why, researchers found that simply getting the basics right – that is, having an appropriate strategy with the right supporting data, technology and skills in place – was not enough.

When you download a dating app, fill out a profile with some of your most private information, and select "allow app to access location" to locate nearby potential love interests, you may feel a little exposed, but you proceed anyway, in order to find those dates. But there is reason to believe that by using these sites, you may be unknowingly submitting to government tracking--and we can't know for sure because of all of the secrecy involved with deals that data brokers make with government agencies. It's yet another demonstration of the need to bring transparency to the data-collection industry. Dating apps ask users for a variety of highly personal information and retain it indefinitely, potentially forever. This can include photos and videos, text conversations with other users, and information on gender, sexual orientation, political affiliation, religion, desire to have children, location, HIV status, and beyond.

Public Service Commission (PSC) put out a job advert for 151 job vacancies and received 12,639 applicants in a span of one month. This does not only happen for government jobs. The high rate of unemployment has led to the mammoth applications for limited job vacancies yet there are about two human resource (HR) personnel to carry out recruitment in a company. In addition, the emergence of the coronavirus has shifted many companies from the traditional to virtual method of recruitment. However, a few questions come to mind; "Are all submitted applications read and scrutinised to the satisfaction of both the job applicant and employer? How long will that process last? Mr Benjamin Lubogo, leader of a recruitment team at Strategic Engagement Company Limited, says it depends on the human resource manager as well as the recruitment system used at the company. There are different ways of recruiting talent for companies including computer aided software, the traditional method of emailing and most recently artificial intelligence. In Uganda, he says, the most common method of recruitment is the traditional emailing method because of its affordability compared to computer aided software and artificial intelligence. With the traditional emailing system, he says, a company receives applications through an email while computer aided software is automated to notify a company whenever an applications is received with the capability to give job applicants feedback on the progress of their application process. We look through them and that is what a normal HR will do. However, we have heard of situations where people beat the process in that if they get three good CVs, why waste time with all the others," he reveals adding that there are many cases where non- technical HR personnel practicing human resources do not understand professional standards about a certain profession.

As teachers and administrators strive to improve student performance and graduation rates, they're increasingly leveraging new Educational Technology (EdTech) to deliver a higher quality learning experience. To gain a competitive advantage, EdTech market players are integrating advanced technologies such as augmented reality (AR), virtual reality (VR), artificial intelligence (AI), robotics, and Blockchain that are set to be the largest revenue contributors to the education sector in the coming years. In the UAE, 1.2 million school and university students started their e-learning journey a year ago with the onset of the pandemic, which has fueled the surge of EdTech startups. The EdTech sector has been gaining significant momentum, leading to an acceleration of investments in 2020. For instance, the regional EdTech companies raised almost $4m in March last year.

The standard problem setting in Dec-POMDPs is self-play, where the goal is to find a set of policies that play optimally together. Policies learned through self-play may adopt arbitrary conventions and rely on multi-step counterfactual reasoning based on assumptions about other agents' actions and thus fail when paired with humans or independently trained agents. In contrast, no current methods can learn optimal policies that are fully grounded, i.e., do not rely on counterfactual information from observing other agents' actions. To address this, we present off-belief learning} (OBL): at each time step OBL agents assume that all past actions were taken by a given, fixed policy ($\pi_0$), but that future actions will be taken by an optimal policy under these same assumptions. When $\pi_0$ is uniform random, OBL learns the optimal grounded policy. OBL can be iterated in a hierarchy, where the optimal policy from one level becomes the input to the next. This introduces counterfactual reasoning in a controlled manner. Unlike independent RL which may converge to any equilibrium policy, OBL converges to a unique policy, making it more suitable for zero-shot coordination. OBL can be scaled to high-dimensional settings with a fictitious transition mechanism and shows strong performance in both a simple toy-setting and the benchmark human-AI/zero-shot coordination problem Hanabi.

Forecasting short-term motion of nearby vehicles presents an inherently challenging issue as the space of their possible future movements is not strictly limited to a set of single trajectories. Recently proposed techniques that demonstrate plausible results concentrate primarily on forecasting a fixed number of deterministic predictions, or on classifying over a wide variety of trajectories that were previously generated using e.g. dynamic model. This paper focuses on addressing the uncertainty associated with the discussed task by utilising the stochastic nature of generative models in order to produce a diverse set of plausible paths with regards to tracked vehicles. More specifically, we propose to account for the multi-modality of the problem with use of Conditional Variational Autoencoder (C-VAE) conditioned on an agent's past motion as well as a rasterised scene context encoded with Capsule Network (CapsNet). In addition, we demonstrate advantages of employing the Minimum over N (MoN) cost function which measures the distance between ground truth and N generated samples and tries to minimise the loss with respect to the closest sample, effectively leading to more diverse predictions. We examine our network on a publicly available dataset against recent state-of-the-art methods and show that our approach outperforms these techniques in numerous scenarios whilst significantly reducing the number of trainable parameters as well as allowing to sample an arbitrary amount of diverse trajectories.

As urban environments manifest high levels of complexity it is of vital importance that safety systems embedded within autonomous vehicles (AVs) are able to accurately anticipate short-term future motion of nearby agents. This problem can be further understood as generating a sequence of coordinates describing the future motion of the tracked agent. Various proposed approaches demonstrate significant benefits of using a rasterised top-down image of the road, with a combination of Convolutional Neural Networks (CNNs), for extraction of relevant features that define the road structure (eg. driveable areas, lanes, walkways). In contrast, this paper explores use of Capsule Networks (CapsNets) in the context of learning a hierarchical representation of sparse semantic layers corresponding to small regions of the High-Definition (HD) map. Each region of the map is dismantled into separate geometrical layers that are extracted with respect to the agent's current position. By using an architecture based on CapsNets the model is able to retain hierarchical relationships between detected features within images whilst also preventing loss of spatial data often caused by the pooling operation. We train and evaluate our model on publicly available dataset nuTonomy scenes and compare it to recently published methods. We show that our model achieves significant improvement over recently published works on deterministic prediction, whilst drastically reducing the overall size of the network.