In membership/subscriber acquisition and retention, we sometimes need to recommend marketing content for multiple pages in sequence. Different from general sequential decision making process, the use cases have a simpler flow where customers per seeing recommended content on each page can only return feedback as moving forward in the process or dropping from it until a termination state. We refer to this type of problems as sequential decision making in linear--flow. We propose to formulate the problem as an MDP with Bandits where Bandits are employed to model the transition probability matrix. At recommendation time, we use Thompson sampling (TS) to sample the transition probabilities and allocate the best series of actions with analytical solution through exact dynamic programming. The way that we formulate the problem allows us to leverage TS's efficiency in balancing exploration and exploitation and Bandit's convenience in modeling actions' incompatibility. In the simulation study, we observe the proposed MDP with Bandits algorithm outperforms Q-learning with $\epsilon$-greedy and decreasing $\epsilon$, independent Bandits, and interaction Bandits. We also find the proposed algorithm's performance is the most robust to changes in the across-page interdependence strength.

In this thesis, we explore and apply methods inspired by the free energy principle to two important areas in machine learning and neuroscience. The free energy principle is a general mathematical theory of the necessary information-theoretic behaviours of systems which maintain a separation from their environment. A core postulate of the theory is that complex systems can be seen as performing variational Bayesian inference and minimizing an information-theoretic quantity called the variational free energy. The free energy principle originated in, and has been extremely influential in theoretical neuroscience, having spawned a number of neurophysiologically realistic process theories, and maintaining close links with Bayesian Brain viewpoints. The thesis is split into three main parts where we apply methods and insights from the free energy principle to understand questions first in perception, then action, and finally learning. Specifically, in the first section, we focus on the theory of predictive coding, a neurobiologically plausible process theory derived from the free energy principle under certain assumptions, which argues that the primary function of the brain is to minimize prediction errors. We focus on scaling up predictive coding architectures and simulate large-scale predictive coding networks for perception on machine learning benchmarks; we investigate predictive coding's relationship to other classical filtering algorithms, and we demonstrate that many biologically implausible aspects of current models of predictive coding can be relaxed without unduly harming the performance of predictive coding models which allows for a potentially more literal translation of predictive coding theory into cortical microcircuits. In the second part of the thesis, we focus on the application of methods deriving from the free energy principle to action. We study the extension of methods of'active inference', a neurobiologically grounded account of action through variational message passing, to utilize deep artificial neural networks, allowing these methods to'scale up' to be competitive with state of the art deep reinforcement learning methods.

Current domain-independent, classical planners require symbolic models of the problem domain and instance as input, resulting in a knowledge acquisition bottleneck. Meanwhile, although deep learning has achieved significant success in many fields, the knowledge is encoded in a subsymbolic representation which is incompatible with symbolic systems such as planners. We propose Latplan, an unsupervised architecture combining deep learning and classical planning. Given only an unlabeled set of image pairs showing a subset of transitions allowed in the environment (training inputs), Latplan learns a complete propositional PDDL action model of the environment. Later, when a pair of images representing the initial and the goal states (planning inputs) is given, Latplan finds a plan to the goal state in a symbolic latent space and returns a visualized plan execution. We evaluate Latplan using image-based versions of 6 planning domains: 8-puzzle, 15-Puzzle, Blocksworld, Sokoban and Two variations of LightsOut.

This paper surveys the field of multiagent deep reinforcement learning. The combination of deep neural networks with reinforcement learning has gained increased traction in recent years and is slowly shifting the focus from single-agent to multiagent environments. Dealing with multiple agents is inherently more complex as (a) the future rewards depend on the joint actions of multiple players and (b) the computational complexity of functions increases. We present the most common multiagent problem representations and their main challenges, and identify five research areas that address one or more of these challenges: centralised training and decentralised execution, opponent modelling, communication, efficient coordination, and reward shaping. We find that many computational studies rely on unrealistic assumptions or are not generalisable to other settings; they struggle to overcome the curse of dimensionality or nonstationarity. Approaches from psychology and sociology capture promising relevant behaviours such as communication and coordination. We suggest that, for multiagent reinforcement learning to be successful, future research addresses these challenges with an interdisciplinary approach to open up new possibilities for more human-oriented solutions in multiagent reinforcement learning.

Current model-based reinforcement learning methods struggle when operating from complex visual scenes due to their inability to prioritize task-relevant features. To mitigate this problem, we propose learning Task Informed Abstractions (TIA) that explicitly separates reward-correlated visual features from distractors. For learning TIA, we introduce the formalism of Task Informed MDP (TiMDP) that is realized by training two models that learn visual features via cooperative reconstruction, but one model is adversarially dissociated from the reward signal. Empirical evaluation shows that TIA leads to significant performance gains over state-of-the-art methods on many visual control tasks where natural and unconstrained visual distractions pose a formidable challenge.

Approaches based on refinement operators have been successfully applied to class expression learning on RDF knowledge graphs. These approaches often need to explore a large number of concepts to find adequate hypotheses. This need arguably stems from current approaches relying on myopic heuristic functions to guide their search through an infinite concept space. In turn, deep reinforcement learning provides effective means to address myopia by estimating how much discounted cumulated future reward states promise. In this work, we leverage deep reinforcement learning to accelerate the learning of concepts in $\mathcal{ALC}$ by proposing DRILL -- a novel class expression learning approach that uses a convolutional deep Q-learning model to steer its search. By virtue of its architecture, DRILL is able to compute the expected discounted cumulated future reward of more than $10^3$ class expressions in a second on standard hardware. We evaluate DRILL on four benchmark datasets against state-of-the-art approaches. Our results suggest that DRILL converges to goal states at least 2.7$\times$ faster than state-of-the-art models on all benchmark datasets. We provide an open-source implementation of our approach, including training and evaluation scripts as well as pre-trained models.

Maintaining the stability of the modern power grid is becoming increasingly difficult due to fluctuating power consumption, unstable power supply coming from renewable energies, and unpredictable accidents such as man-made and natural disasters. As the operation on the power grid must consider its impact on future stability, reinforcement learning (RL) has been employed to provide sequential decision-making in power grid management. However, existing methods have not considered the environmental constraints. As a result, the learned policy has risk of selecting actions that violate the constraints in emergencies, which will escalate the issue of overloaded power lines and lead to large-scale blackouts. In this work, we propose a novel method for this problem, which builds on top of the search-based planning algorithm. At the planning stage, the search space is limited to the action set produced by the policy. The selected action strictly follows the constraints by testing its outcome with the simulation function provided by the system. At the learning stage, to address the problem that gradients cannot be propagated to the policy, we introduce Evolutionary Strategies (ES) with black-box policy optimization to improve the policy directly, maximizing the returns of the long run. In NeurIPS 2020 Learning to Run Power Network (L2RPN) competition, our solution safely managed the power grid and ranked first in both tracks.

We propose a novel algorithm named Expert Q-learning. Expert Q-learning was inspired by Dueling Q-learning and aimed at incorporating the ideas from semi-supervised learning into reinforcement learning through splitting Q-values into state values and action advantages. Different from Generative Adversarial Imitation Learning and Deep Q-Learning from Demonstrations, the offline expert we have used only predicts the value of a state from {-1, 0, 1}, indicating whether this is a bad, neutral or good state. An expert network was designed in addition to the Q-network, which updates each time following the regular offline minibatch update whenever the expert example buffer is not empty. The Q-network plays the role of the advantage function only during the update. Our algorithm also keeps asynchronous copies of the Q-network and expert network, predicting the target values using the same manner as of Double Q-learning. We compared on the game of Othello our algorithm with the state-of-the-art Q-learning algorithm, which was a combination of Double Q-learning and Dueling Q-learning. The results showed that Expert Q-learning was indeed useful and more resistant to the overestimation bias of Q-learning. The baseline Q-learning algorithm exhibited unstable and suboptimal behavior, especially when playing against a stochastic player, whereas Expert Q-learning demonstrated more robust performance with higher scores. Expert Q-learning without using examples has also gained better results than the baseline algorithm when trained and tested against a fixed player. On the other hand, Expert Q-learning without examples cannot win against the baseline Q-learning algorithm in direct game competitions despite the fact that it has also shown the strength of reducing the overestimation bias.

Can we apply massive advancements of Transformer approach with its simplicity and scalability to Reinforcement Learning (RL)? Yes, but for that - one needs to approach RL as a sequence modeling problem. The Decision Transformer does that by abstracting RL as a conditional sequence modeling and using language modeling technique of casual masking of self-attention from GPT/BERT, enabling autoregressive generation of trajectories from the previous tokens in a sequence. The classical RL approach of fitting the value functions, or computing policy gradients (needs live correction; online), has been ditched in favor of masked Transformer yielding optimal actions. The Decision Transformer can match or outperform strong algorithms designed explicitly for offline RL with minimal modifications from standard language modeling architectures.

I recently started an AI-focused educational newsletter, that already has over 80,000 subscribers. TheSequence is a no-BS (meaning no hype, no news etc) ML-oriented newsletter that takes 5 minutes to read. The goal is to keep you up to date with machine learning projects, research papers and concepts. Humans have a great ability to reuse prior knowledge when mastering new skills. Just go through the steps you follow to learn a new cooking recipe or a new workout routine.