Although recurrent neural networks (RNNs) for reinforcement learning (RL) have addressed unique advantages in various aspects, e. g., solving memory-dependent tasks and meta-learning, very few studies have demonstrated how RNNs can solve the problem of hierarchical RL by autonomously developing hierarchical control. In this paper, we propose a novel model-free RL framework called ReMASTER, which combines an off-policy actor-critic algorithm with a multiple timescale stochastic recurrent neural network for solving memory-dependent and hierarchical tasks. We performed experiments using a challenging continuous control task and showed that: (1) Internal representation necessary for achieving hierarchical control autonomously develops through exploratory learning. (2) Stochastic neurons in RNNs enable faster relearning when adapting to a new task which is a recomposition of sub-goals previously learned.

Machine learning can provide efficient solutions to the complex problems encountered in autonomous driving, but ensuring their safety remains a challenge. A number of authors have attempted to address this issue, but there are few publicly-available tools to adequately explore the trade-offs between functionality, scalability, and safety. We thus present WiseMove, a software framework to investigate safe deep reinforcement learning in the context of motion planning for autonomous driving. WiseMove adopts a modular learning architecture that suits our current research questions and can be adapted to new technologies and new questions. We present the details of WiseMove, demonstrate its use on a common traffic scenario, and describe how we use it in our ongoing safe learning research.

Expectation-Maximization (EM) is the fallback method for parameter estimation of hidden (aka latent) variable models. Given the full batch of data, EM forms an upper-bound of the negative log-likelihood of the model at each iteration and then updates to the minimizer of this upper-bound. We introduce a versatile online variant of EM where the data arrives in as a stream. Our motivation is based on the relative entropy divergences between two joint distributions over the hidden and visible variables. We view the EM upper-bound as a Monte Carlo approximation of an expectation and show that the joint relative entropy divergence induces a similar expectation form. As a result, we employ the divergence to the old model as the inertia term to motivate our online EM algorithm. Our motivation is more widely applicable than previous ones and leads to simple online updates for mixture of exponential distributions, hidden Markov models, and the first known online update for Kalman filters. Additionally, the finite sample form of the inertia term lets us derive online updates when there is no closed form solution. Experimentally, sweeping the data with an online update converges much faster than the batch update. Our divergence based methods also lead to a simple way to combine hidden variable models and this immediately gives efficient algorithms for distributed setting.

Abstract--Model-free reinforcement learning has recently been shown to successfully learn navigation policies from raw sensor data. In this work, we address the problem of learning driving policies for an autonomous agent in a high-fidelity simulator. Building upon recent research that applies deep reinforcement learning to navigation problems, we present a modular deep reinforcement learning approach to predict the steering angle of the car from raw images. The control module trained with reinforcement learning takes the latent vector as input to predict the correct steering angle. The experimental results have showed that our method is capable of learning to maneuver the car without any human control signals. I. INTRODUCTION Reinforcement learning (RL) is gaining interest as a promising avenueto training end-to-end autonomous driving policies. These algorithms have recently been shown to solve complex tasks such as navigation from raw vision-sensor modalities. However, training those algorithms require vast amounts of data and interactions with the environment to cover a wide variety of driving scenarios. The collection of such data if even possible is costly and time-consuming.

In the changing tech scenario in India, noted and well-established institutes have now also started to step forward and train students as well as the professionals in artificial intelligence and machine learning. The institutes are providing both the current needs of algorithms and mathematical insights as well as practical experiences. In this article, we list 5 tier-1 institutes that have added courses on artificial intelligence in India. About The Programme: This institute launched a dual degree specialisation in data science as well as in robotics in the year 2018. Any B.Tech student can enroll in this programme based on the CGPA cut-off of 8.0 at the end of the 5th semester.

In this paper, we consider the problem of reducing the bit error rate of flash-based solid state drives (SSDs) when cells are subject to inter-cell interference (ICI). By observing that the outputs of adjacent victim cells can be correlated due to common aggressors, we propose a novel channel model to accurately represent the true flash channel. This model, equivalent to a finite-state Markov channel model, allows the use of the sum-product algorithm to calculate more accurate posterior distributions of individual cell inputs given the joint outputs of victim cells. These posteriors can be easily mapped to the log-likelihood ratios that are passed as inputs to the soft LDPC decoder. When the output is available with high precision, our simulation showed that a significant reduction in the bit-error rate can be obtained, reaching $99.99\%$ reduction compared to current methods, when the diagonal coupling is very strong. In the realistic case of low-precision output, our scheme provides less impressive improvements due to information loss in the process of quantization. To improve the performance of the new detector in the quantized case, we propose a new iterative scheme that alternates multiple times between the detector and the decoder. Our simulations showed that the iterative scheme can significantly improve the bit error rate even in the quantized case.

We present a framework to train a structured prediction model by performing smoothing on the inference algorithm it builds upon. Smoothing overcomes the non-smoothness inherent to the maximum margin structured prediction objective, and paves the way for the use of fast primal gradient-based optimization algorithms. We illustrate the proposed framework by developing a novel primal incremental optimization algorithm for the structural support vector machine. The proposed algorithm blends an extrapolation scheme for acceleration and an adaptive smoothing scheme and builds upon the stochastic variance-reduced gradient algorithm. We establish its worst-case global complexity bound and study several practical variants, including extensions to deep structured prediction. We present experimental results on two real-world problems, namely named entity recognition and visual object localization. The experimental results show that the proposed framework allows us to build upon efficient inference algorithms to develop large-scale optimization algorithms for structured prediction which can achieve competitive performance on the two real-world problems.

A wide class of machine learning algorithms can be reduced to variable elimination on factor graphs. While factor graphs provide a unifying notation for these algorithms, they do not provide a compact way to express repeated structure when compared to plate diagrams for directed graphical models. To exploit efficient tensor algebra in graphs with plates of variables, we generalize undirected factor graphs to plated factor graphs and variable elimination to a tensor variable elimination algorithm that operates directly on plated factor graphs. Moreover, we generalize complexity bounds based on treewidth and characterize the class of plated factor graphs for which inference is tractable. As an application, we integrate tensor variable elimination into the Pyro probabilistic programming language to enable exact inference in discrete latent variable models with repeated structure. We validate our methods with experiments on both directed and undirected graphical models, including applications to polyphonic music modeling, animal movement modeling, and latent sentiment analysis.

Reinforcement learning (RL) agents have traditionally been tasked with maximizing the value function of a Markov decision process (MDP), either in continuous settings, with fixed discount factor $\gamma < 1$, or in episodic settings, with $\gamma = 1$. While this has proven effective for specific tasks with well-defined objectives (e.g., games), it has never been established that fixed discounting is suitable for general purpose use (e.g., as a model of human preferences). This paper characterizes rationality in sequential decision making using a set of seven axioms and arrives at a form of discounting that generalizes traditional fixed discounting. In particular, our framework admits a state-action dependent "discount" factor that is not constrained to be less than 1, so long as there is eventual long run discounting. Although this broadens the range of possible preference structures in continuous settings, we show that there exists a unique "optimizing MDP" with fixed $\gamma < 1$ whose optimal value function matches the true utility of the optimal policy, and we quantify the difference between value and utility for suboptimal policies. Our work can be seen as providing a normative justification for (a slight generalization of) Martha White's RL task formalism (2017) and other recent departures from the traditional RL, and is relevant to task specification in RL, inverse RL and preference-based RL.

Trust-region methods have yielded state-of-the-art results in policy search. A common approach is to use KL-divergence to bound the region of trust resulting in a natural gradient policy update. We show that the natural gradient and trust region optimization are equivalent if we use the natural parameterization of a standard exponential policy distribution in combination with compatible value function approximation. Moreover, we show that standard natural gradient updates may reduce the entropy of the policy according to a wrong schedule leading to premature convergence. To control entropy reduction we introduce a new policy search method called compatible policy search (COPOS) which bounds entropy loss. The experimental results show that COPOS yields state-of-the-art results in challenging continuous control tasks and in discrete partially observable tasks.