Reinforcement learning (RL) requires access to a reward function that incentivizes the right behavior, but these are notoriously hard to specify for complex tasks. Preference-based RL provides an alternative: learning policies using a teacher's preferences without pre-defined rewards, thus overcoming concerns associated with reward engineering. However, it is difficult to quantify the progress in preference-based RL due to the lack of a commonly adopted benchmark. In this paper, we introduce B-Pref: a benchmark specially designed for preference-based RL. A key challenge with such a benchmark is providing the ability to evaluate candidate algorithms quickly, which makes relying on real human input for evaluation prohibitive. At the same time, simulating human input as giving perfect preferences for the ground truth reward function is unrealistic. B-Pref alleviates this by simulating teachers with a wide array of irrationalities, and proposes metrics not solely for performance but also for robustness to these potential irrationalities. We showcase the utility of B-Pref by using it to analyze algorithmic design choices, such as selecting informative queries, for state-of-the-art preference-based RL algorithms. We hope that B-Pref can serve as a common starting point to study preference-based RL more systematically. Source code is available at https://github.com/rll-research/B-Pref.

Machine Learning (ML) is more than just training models, the whole workflow must be considered. Once deployed, a ML model needs to be watched and constantly supervised and debugged to guarantee its validity and robustness in unexpected situations. Debugging in ML aims to identify (and address) the model weaknesses in not trivial contexts. Several techniques have been proposed to identify different types of model weaknesses, such as bias in classification, model decay, adversarial attacks, etc., yet there is not a generic framework that allows them to work in a collaborative, modular, portable, iterative way and, more importantly, flexible enough to allow both human- and machine-driven techniques. In this paper, we propose a novel containerized directed graph framework to support and accelerate end-to-end ML workflow management, supervision, and debugging. The framework allows defining and deploying ML workflows in containers, tracking their metadata, checking their behavior in production, and improving the models by using both learned and human-provided knowledge. We demonstrate these capabilities by integrating in the framework two hybrid systems to detect data drift distribution which identify the samples that are far from the latent space of the original distribution, ask for human intervention, and whether retrain the model or wrap it with a filter to remove the noise of corrupted data at inference time. We test these systems on MNIST-C, CIFAR-10-C, and FashionMNIST-C datasets, obtaining promising accuracy results with the help of human involvement.

Fantasy sports allow fans to manage a team of their favorite athletes and compete with friends. The fantasy platform aligns the real-world statistical performance of athletes to fantasy scoring and has steadily risen in popularity to an estimated 9.1 million players per month with 4.4 billion player card views on the ESPN Fantasy Football platform from 2018-2019. In parallel, the sports media community produces news stories, blogs, forum posts, tweets, videos, podcasts and opinion pieces that are both within and outside the context of fantasy sports. However, human fantasy football players can only analyze an average of 3.9 sources of information. Our work discusses the results of a machine learning pipeline to manage an ESPN Fantasy Football team. The use of trained statistical entity detectors and document2vector models applied to over 100,000 news sources and 2.3 million articles, videos and podcasts each day enables the system to comprehend natural language with an analogy test accuracy of 100% and keyword test accuracy of 80%. Deep learning feedforward neural networks provide player classifications such as if a player will be a bust, boom, play with a hidden injury or play meaningful touches with a cumulative 72% accuracy. Finally, a multiple regression ensemble uses the deep learning output and ESPN projection data to provide a point projection for each of the top 500+ fantasy football players in 2018. The point projection maintained a RMSE of 6.78 points. The best fit probability density function from a set of 24 is selected to visualize score spreads. Within the first 6 weeks of the product launch, the total number of users spent a cumulative time of over 4.6 years viewing our AI insights. The training data for our models was provided by a 2015 to 2016 web archive from Webhose, ESPN statistics, and Rotowire injury reports. We used 2017 fantasy football data as a test set.

In recent years BERT shows apparent advantages and great potential in natural language processing tasks. However, both training and applying BERT requires intensive time and resources for computing contextual language representations, which hinders its universality and applicability. To overcome this bottleneck, we propose a deep bidirectional language model by using window masking mechanism at attention layer. This work computes contextual language representations without random masking as does in BERT and maintains the deep bidirectional architecture like BERT. To compute the same sentence representation, our method shows O(n) complexity less compared to other transformer-based models with O($n^2$). To further demonstrate its superiority, computing context language representations on CPU environments is conducted, by using the embeddings from the proposed method, logistic regression shows much higher accuracy in terms of SMS classification. Moverover, the proposed method also achieves significant higher performance in semantic similarity tasks.

Wildfire forecasting is of paramount importance for disaster risk reduction and environmental sustainability. We approach daily fire danger prediction as a machine learning task, using historical Earth observation data from the last decade to predict next-day's fire danger. To that end, we collect, pre-process and harmonize an open-access datacube, featuring a set of covariates that jointly affect the fire occurrence and spread, such as weather conditions, satellite-derived products, topography features and variables related to human activity. We implement a variety of Deep Learning (DL) models to capture the spatial, temporal or spatio-temporal context and compare them against a Random Forest (RF) baseline. We find that either spatial or temporal context is enough to surpass the RF, while a ConvLSTM that exploits the spatio-temporal context performs best with a test Area Under the Receiver Operating Characteristic of 0.926. Our DL-based proof-of-concept provides national-scale daily fire danger maps at a much higher spatial resolution than existing operational solutions.

Language Models are the underpin of all modern Natural Language Processing (NLP) tasks. The introduction of the Transformers architecture has contributed significantly into making Language Modeling very effective across many NLP task, leading to significant advancements in the field. However, Transformers come with a big computational cost, which grows quadratically with respect to the input length. This presents a challenge as to understand long texts requires a lot of context. In this paper, we propose a Fine-Tuning framework, named CoreLM, that extends the architecture of current Pretrained Language Models so that they incorporate explicit entity information. By introducing entity representations, we make available information outside the contextual space of the model, which results in a better Language Model for a fraction of the computational cost. We implement our approach using GPT2 and compare the fine-tuned model to the original. Our proposed model achieves a lower Perplexity in GUMBY and LAMBDADA datasets when compared to GPT2 and a fine-tuned version of GPT2 without any changes. We also compare the models' performance in terms of Accuracy in LAMBADA and Children's Book Test, with and without the use of model-created coreference annotations.

Model quantization is known as a promising method to compress deep neural networks, especially for inferences on lightweight mobile or edge devices. However, model quantization usually requires access to the original training data to maintain the accuracy of the full-precision models, which is often infeasible in real-world scenarios for security and privacy issues. A popular approach to perform quantization without access to the original data is to use synthetically generated samples, based on batch-normalization statistics or adversarial learning. However, the drawback of such approaches is that they primarily rely on random noise input to the generator to attain diversity of the synthetic samples. We find that this is often insufficient to capture the distribution of the original data, especially around the decision boundaries. To this end, we propose Qimera, a method that uses superposed latent embeddings to generate synthetic boundary supporting samples. For the superposed embeddings to better reflect the original distribution, we also propose using an additional disentanglement mapping layer and extracting information from the full-precision model. The experimental results show that Qimera achieves state-of-the-art performances for various settings on data-free quantization.

Objective: Leveraging machine learning methods, we aim to extract both explicit and implicit cause-effect associations in patient-reported, diabetes-related tweets and provide a tool to better understand opinion, feelings and observations shared within the diabetes online community from a causality perspective. Materials and Methods: More than 30 million diabetes-related tweets in English were collected between April 2017 and January 2021. Deep learning and natural language processing methods were applied to focus on tweets with personal and emotional content. A cause-effect-tweet dataset was manually labeled and used to train 1) a fine-tuned Bertweet model to detect causal sentences containing a causal association 2) a CRF model with BERT based features to extract possible cause-effect associations. Causes and effects were clustered in a semi-supervised approach and visualised in an interactive cause-effect-network. Results: Causal sentences were detected with a recall of 68% in an imbalanced dataset. A CRF model with BERT based features outperformed a fine-tuned BERT model for cause-effect detection with a macro recall of 68%. This led to 96,676 sentences with cause-effect associations. "Diabetes" was identified as the central cluster followed by "Death" and "Insulin". Insulin pricing related causes were frequently associated with "Death". Conclusions: A novel methodology was developed to detect causal sentences and identify both explicit and implicit, single and multi-word cause and corresponding effect as expressed in diabetes-related tweets leveraging BERT-based architectures and visualised as cause-effect-network. Extracting causal associations on real-life, patient reported outcomes in social media data provides a useful complementary source of information in diabetes research.

The data center of tomorrow is a data center made up of heterogeneous systems, which will run heterogeneous workloads. The systems will be located as close as possible to the data. Heterogeneous systems will be equipped with binary, biological inspired and quantum accelerators. These architectures will be the foundations to address challenges. Like an orchestra conductor, the hybrid cloud will make it possible to set these systems to music thanks to a layer of security and intelligent automation.

Microsoft is reportedly firming up plans to offer OpenAI's Artificial intelligence (AI) language model GPT-3 to its business customers. The new Azure OpenAI Service will be offered as part of its growing portfolio of business-oriented tools powered by its cloud computing platform. Reporting on the development, The Verge explains that GPT-3 is touted as a new generation of AI language models, which are primarily deployed as autocomplete tools. Microsoft however intends to capitalize on GPT-3's ability to parse language for business tasks such as summarizing documents, analyzing the sentiment of text, and generating ideas for projects and stories. GPT-derived tools aren't exclusive to Microsoft, and The Verge says that startups like Copy.ai use the system to help customers write better work emails and pitch decks.