The AWS AI/ML Community has compiled a series of session guides to AWS re:Invent 2021 to help you get the most out of re:Invent this year. They covered four distinct categories relevant to AI/ML. With a number of our guide authors attending re:Invent virtually, you will find a balance between virtually accessible sessions and sessions available in-person. The AWS AI/ML Community is a vibrant group of developers, data scientists, researchers, and business decision-makers that dive deep into artificial intelligence and machine learning (ML) concepts, contribute with real-world experiences, and collaborate on building projects together. In this video--which should be required viewing for anyone new to re:Invent--Mike dives deep, beyond simply recommending sessions, with loads of tips and advice for how to make the most of your re:Invent experience--in-person or virtual.

AI leads to'great return on investment' in dealing with pandemic The UAE approached the COVID-19 pandemic "as a scientist," said Omar Al Olama, the UAE's Minister of Artificial Intelligence, Digital Economy and Remote Work Applications. Al Olama was appointed by the UAE as the first artificial intelligence (AI) minister in the world in 2017, when he was just 27 years old. That year, his ministry launched a strategy "to become one of the world leaders in AI by 2031." The COVID-19 pandemic, it turns out, may have accelerated the UAE's applications of AI to governance and public health, and to establishing the Emirates as a world leader in AI, as Al Olama, now 31, explained in an exclusive Zoom interview with Al-Monitor on Nov. 18. Al Olama describes a policy response to the pandemic by the UAE that has been data- and analytics-driven and characterized by openness to different ideas, nimbleness in response to changing events, and willingness to accept calculated risks. "We actually were very open to many different solutions, and many different theories out there," he said. "And we worked with everyone, from the East and the West, to try to find the right solutions that can be deployed in the UAE to make us go back to living a relatively normal life. Not the normal life that we're used to. People still need to wear masks. There's still a lot of focus on the general community's safety, but AI played a big role in getting us to this point."

The construction sector constitutes one of the most important sectors in the economy of any country. Many construction projects experience time and cost overruns due to the wrong choice of contractors. In this paper, the feedback integrated multi-criteria group decision support model for contractor selection was proposed. The proposed model consists of two modules; technical evaluation module and financial evaluation module. The technical evaluation module is employed to screen out the contractors to a smaller set of acceptable contractors and the functionality of the module is based on the Fuzzy Analytic Hierarchy Process (FAHP).

In diploid organisms, such as humans, each individual's genome is organized into pairs of chromosomes, each half inherited from each parent. When an individual is an offspring of biologically related parents, both chromosomes of the same pair can share identical segments, creating long stretches of consecutive homozygosity, known as runs of homozygosity (ROH). In the last decades, studies on the identification of ROH carried out in human populations have revealed the presence of ROH even in cosmopolitan non-inbred populations, disclosing an increment of inbreeding levels and the consequent reduction of genetic diversity of populations, which is proportional to the walking distance from Africa, as expected by the out-of-Africa model of human colonization (Ceballos et al., 2018; Kirin et al., 2010; Lemes et al., 2018; Leutenegger et al., 2011; Pemberton et al., 2012). The distribution of ROH along the chromosomes is very uneven, resulting in some genomic regions having significant absence (coldspots) or excess of ROH (ROH islands) (Ceballos et al., 2018). The mechanisms for the emergence of these regions are still under discussion. For example, there is evidence that ROH islands could represent regions that harbor genes target of positive selection since low-recombination regions commonly are locations of selective sweeps, in which a new beneficial mutation increases in frequency and becomes fixed, causing the overall reduction in genetic diversity of the region (Ceballos et al., 2018; Pemberton et al., 2012). To detect ROH and ROH islands, the genetic material of individuals from a given population is genotyped, and a set of single nucleotide polymorphisms (SNPs) is obtained. Each SNP entry is codified to 1 if that SNP belongs to an ROH for that individual and to 0 otherwise, where a marker is defined as belonging to an ROH for an individual if it is surrounded by a region with high frequency of homozygous SNPs.

Algorithmic decisions made by machine learning models in high-stakes domains may have lasting impacts over time. Unfortunately, naive applications of standard fairness criterion in static settings over temporal domains may lead to delayed and adverse effects. To understand the dynamics of performance disparity, we study a fairness problem in Markov decision processes (MDPs). Specifically, we propose return parity, a fairness notion that requires MDPs from different demographic groups that share the same state and action spaces to achieve approximately the same expected time-discounted rewards. We first provide a decomposition theorem for return disparity, which decomposes the return disparity of any two MDPs into the distance between group-wise reward functions, the discrepancy of group policies, and the discrepancy between state visitation distributions induced by the group policies. Motivated by our decomposition theorem, we propose algorithms to mitigate return disparity via learning a shared group policy with state visitation distributional alignment using integral probability metrics. We conduct experiments to corroborate our results, showing that the proposed algorithm can successfully close the disparity gap while maintaining the performance of policies on two real-world recommender system benchmark datasets.

The efficacy of machine learning (ML) models depends on both algorithms and data. Training data defines what we want our models to learn, and testing data provides the means by which their empirical progress is measured. Benchmark datasets define the entire world within which models exist and operate, yet research continues to focus on critiquing and improving the algorithmic aspect of the models rather than critiquing and improving the data with which our models operate. If "data is the new oil," we are still missing work on the refineries by which the data itself could be optimized for more effective use.

Medical Visual Question Answering (VQA) is a combination of medical artificial intelligence and popular VQA challenges. Given a medical image and a clinically relevant question in natural language, the medical VQA system is expected to predict a plausible and convincing answer. Although the general-domain VQA has been extensively studied, the medical VQA still needs specific investigation and exploration due to its task features. In the first part of this survey, we cover and discuss the publicly available medical VQA datasets up to date about the data source, data quantity, and task feature. In the second part, we review the approaches used in medical VQA tasks. In the last part, we analyze some medical-specific challenges for the field and discuss future research directions.

Minimizers and convolutional neural networks (CNNs) are two quite distinct popular techniques that have both been employed to analyze biological sequences. At face value, the methods seem entirely dissimilar. Minimizers use min-wise hashing on a rolling window to extract a single important k-mer feature per window. CNNs start with a wide array of randomly initialized convolutional filters, paired with a pooling operation, and then multiple additional neural layers to learn both the filters themselves and how those filters can be used to classify the sequence. In this manuscript, I demonstrate through a careful mathematical analysis of hash function properties that there are deep theoretical connections between minimizers and convolutional filters -- in short, for sequences over a categorical alphabet, random Gaussian initialization of convolutional filters with max-pooling is equivalent to choosing minimizers from a random hash function biased towards more distinct k-mers. This provides a partial explanation for the unreasonable effectiveness of CNNs in categorical sequence analysis.

IndyGeneUS AI (pronounced "indigenous", a South Africa Founder Institute portfolio company) opens its seed round with a $1.5M investment from South African based venture capital company IsimoVest Venture Capital Partners in addition to an undisclosed amount from a South African angel investor. Located in Cape Town, IsimoVest is a Pan-African venture capital company that specializes in early stage, high-potential technology ventures that will likely provide highly impactful corporate social investment solutions. As a cutting-edge investment firm, IsimoVest is dedicated to economically empowering communities for the benefit of Africans. To that end, the firm will support IndyGeneUS AI in establishing its sequencing facility in Cape Town and establishing local key strategic partnerships. IndyGeneUS AI is developing a Multiomic data analysis and management platform that can detect new signature sequences such as biomarkers or polygenic risk scores by integrating "omics" data, meta data and textual information such as Electronic Healthcare Record data.

Precision medicine requires big data. In order to improve the treatment of individuals with cancer, or to understand rare diseases, scientists and clinicians, as well as AI technologies require access to larger sets of health research data that covers diverse populations and wide ranges of conditions. For AI, more data means a better understanding of diseases, which will lead to more accurate diagnosis and treatment. At the same time, each hospital will only see a relatively small number of individuals with a disease, and even across the province, we have access to only a small portion of the total data available worldwide. To build the large-scale datasets needed to drive forward precision medicine, sharing of data across the country and around the world is critical.