Recommendation systems play an important role in today's digital world. They have found applications in various applications such as music platforms, e.g., Spotify, and movie streaming services, e.g., Netflix. Less research effort has been devoted to physical exercise recommendation systems. Sedentary lifestyles have become the major driver of several diseases as well as healthcare costs. In this paper, we develop a recommendation system for daily exercise activities to users based on their history, profile and similar users. The developed recommendation system uses a deep recurrent neural network with user-profile attention and temporal attention mechanisms. Moreover, exercise recommendation systems are significantly different from streaming recommendation systems in that we are not able to collect click feedback from the participants in exercise recommendation systems. Thus, we propose a real-time, expert-in-the-loop active learning procedure. The active learners calculate the uncertainty of the recommender at each time step for each user and ask an expert for a recommendation when the certainty is low. In this paper, we derive the probability distribution function of marginal distance, and use it to determine when to ask experts for feedback. Our experimental results on a mHealth dataset show improved accuracy after incorporating the real-time active learner with the recommendation system.

Semi-supervised learning (SSL) has demonstrated its potential to improve the model accuracy for a variety of learning tasks when the high-quality supervised data is severely limited. Although it is often established that the average accuracy for the entire population of data is improved, it is unclear how SSL fares with different sub-populations. Understanding the above question has substantial fairness implications when these different sub-populations are defined by the demographic groups we aim to treat fairly. In this paper, we reveal the disparate impacts of deploying SSL: the sub-population who has a higher baseline accuracy without using SSL (the ``rich" sub-population) tends to benefit more from SSL; while the sub-population who suffers from a low baseline accuracy (the ``poor" sub-population) might even observe a performance drop after adding the SSL module. We theoretically and empirically establish the above observation for a broad family of SSL algorithms, which either explicitly or implicitly use an auxiliary ``pseudo-label". Our experiments on a set of image and text classification tasks confirm our claims. We discuss how this disparate impact can be mitigated and hope that our paper will alarm the potential pitfall of using SSL and encourage a multifaceted evaluation of future SSL algorithms. Code is available at github.com/UCSC-REAL/Disparate-SSL.

Causal discovery, i.e., inferring underlying cause-effect relationships from observations of a scene or system, is an inherent mechanism in human cognition, but has been shown to be highly challenging to automate. The majority of approaches in the literature aiming for this task consider constrained scenarios with fully observed variables or data from stationary time-series. In this work we aim for causal discovery in a more general class of scenarios, scenes with non-stationary behavior over time. For our purposes we here regard a scene as a composition objects interacting with each other over time. Non-stationarity is modeled as stationarity conditioned on an underlying variable, a state, which can be of varying dimension, more or less hidden given observations of the scene, and also depend more or less directly on these observations. We propose a probabilistic deep learning approach called State-Dependent Causal Inference (SDCI) for causal discovery in such conditionally stationary time-series data. Results in two different synthetic scenarios show that this method is able to recover the underlying causal dependencies with high accuracy even in cases with hidden states.

Solutions to the Traveling Salesperson Problem (TSP) have practical applications to processes in transportation, logistics, and automation, yet must be computed with minimal delay to satisfy the real-time nature of the underlying tasks. However, solving large TSP instances quickly without sacrificing solution quality remains challenging for current approximate algorithms. To close this gap, we present a hybrid data-driven approach for solving the TSP based on Graph Neural Networks (GNNs) and Guided Local Search (GLS). Our model predicts the regret of including each edge of the problem graph in the solution; GLS uses these predictions in conjunction with the original problem graph to find solutions. Our experiments demonstrate that this approach converges to optimal solutions at a faster rate than state-of-the-art learning-based approaches and non-learning GLS algorithms for the TSP, notably finding optimal solutions to 96% of the 50-node problem set, 7% more than the next best benchmark, and to 20% of the 100-node problem set, 4.5x more than the next best benchmark. When generalizing from 20-node problems to the 100-node problem set, our approach finds solutions with an average optimality gap of 2.5%, a 10x improvement over the next best learning-based benchmark.

In this post, I will explore how the advent of large pre-trained language models (such as GPT3 [1]) are giving rise to the new paradigm of'prompt engineering' in the field of NLP. This new paradigm allows us to rapidly prototype complex NLP applications with little to no effort and based on very small amounts of data. I will present a case study where I used this technique during my summer internship at Waylay to create an application that makes industry-level automatization accessible to everyone using voice and text inputs (think of something like google assistant, but for IoT and on steroids!). Finally, I will conclude with some remarks on this new and exciting trend. If you don't feel like reading, you can watch this recording of the internal meeting where I presented my solution to the company.

Synthetic media, better known as deepfakes, could be a goldmine for filmmakers. But the technology has already terrorized women who have had their faces inserted into pornography. And it could potentially disrupt society. You may never have heard the term « synthetic media »-- more commonly known as « deepfakes »-- but our military, law enforcement and intelligence agencies certainly have. They are hyper-realistic video and audio recordings that use artificial intelligence and « deep » learning to create « fake » content or « deepfakes.

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. DeepMind, the U.K.-based AI lab that seeks to develop artificial general intelligence, has finally become profitable, according to the company's latest financial report. Since being acquired by Google (now Alphabet Inc.) in 2014, DeepMind has struggled to break even with its growing expenses. And now, it is finally giving its parent company and shareholders hopeful signs that it has earned its place among Alphabet's constellation of profitable businesses. This could be wonderful news for the AI lab, which has been hemorrhaging large sums throughout its entire life.

According to a corporate filing with the UK company registry published on Tuesday, Google-backed artificial intelligence research firm DeepMind turned a profit for the first time last year and saw a significant rise in its revenues. As per the filings, its sales rose by £560 million last year, touching £826 million, compared to £266 million in 2019. The surge in revenue helped the company turn its first profit of £43.8 million, compared to a loss of £649 million the previous year. The research lab draws its revenue from research and development carried out for other companies under the Alphabet umbrella, including Google, YouTube and X, the moonshot division.

In the machine learning world, the sizes of artificial neural networks -- and their outsize successes -- are creating conceptual conundrums. When a network named AlexNet won an annual image recognition competition in 2012, it had about 60 million parameters. These parameters, fine-tuned during training, allowed AlexNet to recognize images that it had never seen before. Two years later, a network named VGG wowed the competition with more than 130 million such parameters. Some artificial neural networks, or ANNs, now have billions of parameters. These massive networks -- astoundingly successful at tasks such as classifying images, recognizing speech and translating text from one language to another -- have begun to dominate machine learning and artificial intelligence.

In the field of deep learning one subject of research that is emerging rapidly is dynamic neural networks. When we talk about traditional static neural networks we train them with fixed parameters and fix problem-solving skills. But it is well known that the attributes of the input and the environments are changing rapidly in these changing scenarios. So we need something which can change itself automatically according to the input and environment. Here dynamic neural networks are the models which are made with their adapting nature.