Next Words Prediction Using Recurrent NeuralNetworks

Recurrent Partial Words

Partial words are sequences over a finite alphabet that may contain wildcard symbols, called holes, which match or are compatible with all letters; partial words without holes are said to be full words (or simply words). Given an infinite partial word w, the number of distinct full words over the alphabet that are compatible with factors of w of length n, called subwords of w, refers to a measu...

Logfile Failure Prediction using Recurrent and Quasi Recurrent Neural Networks

Inumerable man hours are spent waiting for running programs to finish and then debugging them when they crash by analyzing the logfile. Often times logfiles will contain a plethora of text extraneous to debugging, providing for a tedious undertaking. Therefore, we present an application of state-of-the-art natural language processing with neural networks to classify a run’s final outcome prior ...

Using Recurrent Neural Networks in Trajectory Prediction

MIMO Channel Prediction Using Recurrent Neural Networks

Adaptive modulation is a communication technique capable of maximizing throughput while guaranteeing a fixed symbol error rate (SER). However, this technique requires instantaneous channel state information at the transmitter. This can be obtained by predicting channel states at the receiver and feeding them back to the trasnmitter. Existing algorithms used to predict single-input single-output...

Using the words toddlers know now to predict the words they will learn next

We set forth to show that lexical connectivity plays a role in understanding early word learning. By considering words that are learned in temporal proximity to one another to be related, we are able to better predict the words next learned by toddlers. We build conditional probability models based on data from the growing vocabularies of 77 toddlers, followed longitudinally for a year. This ty...