Many sequential processing tasks require complex nonlinear transition functions from one step to the next. However, recurrent neural networks with “deep” transition functions remain difficult to train, even when using Long Short-Term Memory (LSTM) networks. We introduce a novel theoretical analysis of recurrent networks based on Gersgorin’s circle theorem that illuminates several modeling and optimization issues and improves our understanding of the LSTM cell. Based on this analysis we propose Recurrent Highway Networks, which extend the LSTM architecture to allow step-to-step transition depths larger than one. Several language modeling experiments demonstrate that the proposed architecture results in powerful and efficient models. On the Penn Treebank corpus, solely increasing the transition depth from 1 to 10 improves word-level perplexity from 90.6 to 65.4 using the same number of parameters. On the larger Wikipedia datasets for character prediction (text8 and enwik8), RHNs outperform all previous results and achieve an entropy of 1.27 bits per character.

Several variants of the long short-term memory (LSTM) architecture for recurrent neural networks have been proposed since its inception in 1995. In recent years, these networks have become the state-of-the-art models for a variety of machine learning problems. This has led to a renewed interest in understanding the role and utility of various computational components of typical LSTM variants. In this paper, we present the first large-scale analysis of eight LSTM variants on three representative tasks: speech recognition, handwriting recognition, and polyphonic music modeling. The hyperparameters of all LSTM variants for each task were optimized separately using random search, and their importance was assessed using the powerful functional ANalysis Of VAriance framework. In total, we summarize the results of 5400 experimental runs ( ≈ 15 years of CPU time), which makes our study the largest of its kind on LSTM networks. Our results show that none of the variants can improve upon the standard LSTM architecture significantly, and demonstrate the forget gate and the output activation function to be its most critical components. We further observe that the studied hyperparameters are virtually independent and derive guidelines for their efficient adjustment.

A new indirect scheme for encoding neural network connection weights as sets of wavelet-domain coefficients is proposed in this paper. It exploits spatial regularities in the weight-space to reduce the gene-space dimension by considering the low-frequency wavelet coefficients only. The wavelet-based encoding builds on top of a frequency-domain encoding, but unlike when using a Fourier-type transform, it offers gene locality while preserving continuity of the genotype-phenotype mapping. We argue that this added property allows for more efficient evolutionary search and demonstrate this on the octopus-arm control task, where superior solutions were found in fewer generations. The scalability of the wavelet-based encoding is shown by evolving networks with many parameters to control game-playing agents in the Arcade Learning Environment.

Lipreading, i.e. speech recognition from visual-only recordings of a speaker's face, can be achieved with a processing pipeline based solely on neural networks, yielding significantly better accuracy than conventional methods. Feedforward and recurrent neural network layers (namely Long Short-Term Memory; LSTM) are stacked to form a single structure which is trained by back-propagating error gradients through all the layers. The performance of such a stacked network was experimentally evaluated and compared to a standard Support Vector Machine classifier using conventional computer vision features (Eigenlips and Histograms of Oriented Gradients). The evaluation was performed on data from 19 speakers of the publicly available GRID corpus. With 51 different words to classify, we report a best word accuracy on held-out evaluation speakers of 79.6% using the end-to-end neural network-based solution (11.6% improvement over the best feature-based solution evaluated).

Dealing with high-dimensional input spaces, like visual input, is a challenging task for reinforcement learning (RL). Neuroevolution (NE), used for continuous RL problems, has to either reduce the problem dimensionality by (1) compressing the representation of the neural network controllers or (2) employing a pre-processor (compressor) that transforms the high-dimensional raw inputs into low-dimensional features. In this paper, we are able to evolve extremely small recurrent neural network (RNN) controllers for a task that previously required networks with over a million weights. The high-dimensional visual input, which the controller would normally receive, is first transformed into a compact feature vector through a deep, max-pooling convolutional neural network (MPCNN). Both the MPCNN preprocessor and the RNN controller are evolved successfully to control a car in the TORCS racing simulator using only visual input. This is the first use of deep learning in the context evolutionary RL.

Sequence prediction and classification are ubiquitous and challenging problems in machine learning that can require identifying complex dependencies between temporally distant inputs. Recurrent Neural Networks (RNNs) have the ability, in theory, to cope with these temporal dependencies by virtue of the short-term memory implemented by their recurrent (feedback) connections. However, in practice they are difficult to train successfully when long-term memory is required. This paper introduces a simple, yet powerful modification to the simple RNN (SRN) architecture, the Clockwork RNN (CW-RNN), in which the hidden layer is partitioned into separate modules, each processing inputs at its own temporal granularity, making computations only at its prescribed clock rate. Rather than making the standard RNN models more complex, CW-RNN reduces the number of SRN parameters, improves the performance significantly in the tasks tested, and speeds up the network evaluation. The network is demonstrated in preliminary experiments involving three tasks: audio signal generation, TIMIT spoken word classification, where it outperforms both SRN and LSTM networks, and online handwriting recognition, where it outperforms SRNs.

The idea of using evolutionary computation to train artificial neural networks, or neuroevolution (NE), for reinforcement learning (RL) tasks has now been around for over 20 years. However, as RL tasks become more challenging, the networks required become larger, as do their genomes. But, scaling NE to large nets (i.e. tens of thousands of weights) is infeasible using direct encodings that map genes one-to-one to network components. In this paper, we scale-up our compressed network encoding where network weight matrices are represented indirectly as a set of Fourier-type coefficients, to tasks that require very-large networks due to the high-dimensionality of their input space. The approach is demonstrated successfully on two reinforcement learning tasks in which the control networks receive visual input: (1) a vision-based version of the octopus control task requiring networks with over 3 thousand weights, and (2) a version of the TORCS driving game where networks with over 1 million weights are evolved to drive a car around a track using video images from the driver's perspective.

We propose a new indirect encoding scheme for neural networks in which the weight matrices are represented in the frequency domain by sets Fourier coefficients. This scheme exploits spatial regularities in the matrix to reduce the dimensionality of the representation by ignoring high-frequency coefficients, as is done in lossy image compression. We compare the efficiency of searching in this "compressed" network space to searching in the space of directly encoded networks, using the CoSyNE neuroevolution algorithm on three benchmark problems: pole-balancing, ball throwing and octopus arm control. The results show that this encoding can dramatically reduce the search space dimensionality such that solutions can be found in significantly fewer evaluations

In this paper we describe simulation of autonomous robots controlled by recurrent neural networks, which are evolved through indirect encoding using HyperNEAT algorithm. The robots utilize 180 degree wide sensor array. Thanks to the scalability of the neural network generated by HyperNEAT, the sensor array can have various resolution. This would allow to use camera as an input for neural network controller used in real robot. The robots were simulated using software simulation environment. In the experiments the robots were trained to drive with imaximum average speed. Such fitness forces them to learn how to drive on roads and avoid collisions. Evolved neural networks show excellent scalability. Scaling of the sensory input breaks performance of the robots, which should be gained back with re-training of the robot with a different sensory input resolution.

Modifications of an existing neural network called categorizing and learning module (CALM) that allow learning of temporal sequences are introduced in this paper. We embedded an associative learning mechanism which allows to look into the past when classifying present stimuli. We have built in the Euclidean metrics instead of the weighted sum found in the original learning rule. This improvement allows better discrimination in case of learning low dimensional patterns in the temporal sequences. Results were obtained from testing the enhanced module on simple artificial data. These experiments promise applicability of the enhanced module in a real problem domain.

In this paper an enhancement of the categorizing and learning module (CALM) is presented. The CALM neural network module was used as a basic building block of the more complex neural network for vision. The CALM module stands there as a model of cortical columns, which compose the primary visual cortex. The main disadvantage of the basic CALM model found in the previous research is that the module is not capable to detect a change of input signal automatically, which cannot be satisfactory from the biological point of view. We describe certain modifications of the structure and behavior of the module that make the model more autonomous and suitable for the modeling of the human visual cortex.